id
large_stringlengths 9
16
| submitter
large_stringlengths 3
64
⌀ | authors
large_stringlengths 5
20.2k
| title
large_stringlengths 5
245
| comments
large_stringlengths 1
482
⌀ | journal-ref
large_stringlengths 4
382
⌀ | doi
large_stringlengths 9
151
⌀ | report-no
large_stringlengths 2
509
⌀ | categories
large_stringlengths 5
78
| license
large_stringclasses 9
values | abstract
large_stringlengths 83
3.55k
| versions
large listlengths 1
21
| update_date
timestamp[ms]date 2007-05-23 00:00:00
2025-05-30 00:00:00
| authors_parsed
large listlengths 1
1.35k
| prompt
large_stringlengths 166
3.64k
| label
stringclasses 2
values | prob
float64 0.5
0.73
|
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
2307.02057 | Markus Bause | Mathias Anselmann, Markus Bause, Nils Margenberg, Pavel Shamko | Benchmark computations of dynamic poroelasticity | arXiv admin note: text overlap with arXiv:2303.06742 | null | null | null | math.NA cs.NA | http://creativecommons.org/licenses/by-nc-nd/4.0/ | We present benchmark computations of dynamic poroelasticity modeling fluid
flow in deformable porous media by a coupled hyperbolic-parabolic system of
partial differential equations. A challenging benchmark setting and goal
quantities of physical interest for this problem are proposed. Computations
performed by space-time finite element approximations with continuous and
discontinuous discretizations of the time variable are summarized. By this work
we intend to stimulate comparative studies by other research groups for the
evaluation of dynamic poroelasticity solver regarding the accuracy of
discretization techniques, the efficiency and robustness of iterative methods
for the linear systems and the arrangement of the model equations in terms of
their variables (two-field or multi-field formulations).
| [
{
"version": "v1",
"created": "Wed, 5 Jul 2023 06:47:31 GMT"
}
] | 2023-07-06T00:00:00 | [
[
"Anselmann",
"Mathias",
""
],
[
"Bause",
"Markus",
""
],
[
"Margenberg",
"Nils",
""
],
[
"Shamko",
"Pavel",
""
]
] | TITLE: Benchmark computations of dynamic poroelasticity
ABSTRACT: We present benchmark computations of dynamic poroelasticity modeling fluid
flow in deformable porous media by a coupled hyperbolic-parabolic system of
partial differential equations. A challenging benchmark setting and goal
quantities of physical interest for this problem are proposed. Computations
performed by space-time finite element approximations with continuous and
discontinuous discretizations of the time variable are summarized. By this work
we intend to stimulate comparative studies by other research groups for the
evaluation of dynamic poroelasticity solver regarding the accuracy of
discretization techniques, the efficiency and robustness of iterative methods
for the linear systems and the arrangement of the model equations in terms of
their variables (two-field or multi-field formulations).
| no_new_dataset | 0.70825 |
2307.02065 | Omar Moured | Omar Moured, Jiaming Zhang, Alina Roitberg, Thorsten Schwarz, Rainer
Stiefelhagen | Line Graphics Digitization: A Step Towards Full Automation | Accepted at The 17th International Conference on Document Analysis
and Recognition (ICDAR 2023) | null | null | null | cs.CV cs.AI cs.LG | http://creativecommons.org/licenses/by-nc-nd/4.0/ | The digitization of documents allows for wider accessibility and
reproducibility. While automatic digitization of document layout and text
content has been a long-standing focus of research, this problem in regard to
graphical elements, such as statistical plots, has been under-explored. In this
paper, we introduce the task of fine-grained visual understanding of
mathematical graphics and present the Line Graphics (LG) dataset, which
includes pixel-wise annotations of 5 coarse and 10 fine-grained categories. Our
dataset covers 520 images of mathematical graphics collected from 450 documents
from different disciplines. Our proposed dataset can support two different
computer vision tasks, i.e., semantic segmentation and object detection. To
benchmark our LG dataset, we explore 7 state-of-the-art models. To foster
further research on the digitization of statistical graphs, we will make the
dataset, code, and models publicly available to the community.
| [
{
"version": "v1",
"created": "Wed, 5 Jul 2023 07:08:58 GMT"
}
] | 2023-07-06T00:00:00 | [
[
"Moured",
"Omar",
""
],
[
"Zhang",
"Jiaming",
""
],
[
"Roitberg",
"Alina",
""
],
[
"Schwarz",
"Thorsten",
""
],
[
"Stiefelhagen",
"Rainer",
""
]
] | TITLE: Line Graphics Digitization: A Step Towards Full Automation
ABSTRACT: The digitization of documents allows for wider accessibility and
reproducibility. While automatic digitization of document layout and text
content has been a long-standing focus of research, this problem in regard to
graphical elements, such as statistical plots, has been under-explored. In this
paper, we introduce the task of fine-grained visual understanding of
mathematical graphics and present the Line Graphics (LG) dataset, which
includes pixel-wise annotations of 5 coarse and 10 fine-grained categories. Our
dataset covers 520 images of mathematical graphics collected from 450 documents
from different disciplines. Our proposed dataset can support two different
computer vision tasks, i.e., semantic segmentation and object detection. To
benchmark our LG dataset, we explore 7 state-of-the-art models. To foster
further research on the digitization of statistical graphs, we will make the
dataset, code, and models publicly available to the community.
| new_dataset | 0.715629 |
2307.02078 | Zheheng Luo | Zheheng Luo, Lei Liu, Qianqian Xie, Sophia Ananiadou | Graph Contrastive Topic Model | 17 pages, 4 figures | null | null | null | cs.CL | http://creativecommons.org/licenses/by/4.0/ | Existing NTMs with contrastive learning suffer from the sample bias problem
owing to the word frequency-based sampling strategy, which may result in false
negative samples with similar semantics to the prototypes. In this paper, we
aim to explore the efficient sampling strategy and contrastive learning in NTMs
to address the aforementioned issue. We propose a new sampling assumption that
negative samples should contain words that are semantically irrelevant to the
prototype. Based on it, we propose the graph contrastive topic model (GCTM),
which conducts graph contrastive learning (GCL) using informative positive and
negative samples that are generated by the graph-based sampling strategy
leveraging in-depth correlation and irrelevance among documents and words. In
GCTM, we first model the input document as the document word bipartite graph
(DWBG), and construct positive and negative word co-occurrence graphs (WCGs),
encoded by graph neural networks, to express in-depth semantic correlation and
irrelevance among words. Based on the DWBG and WCGs, we design the
document-word information propagation (DWIP) process to perform the edge
perturbation of DWBG, based on multi-hop correlations/irrelevance among
documents and words. This yields the desired negative and positive samples,
which will be utilized for GCL together with the prototypes to improve learning
document topic representations and latent topics. We further show that GCL can
be interpreted as the structured variational graph auto-encoder which maximizes
the mutual information of latent topic representations of different
perspectives on DWBG. Experiments on several benchmark datasets demonstrate the
effectiveness of our method for topic coherence and document representation
learning compared with existing SOTA methods.
| [
{
"version": "v1",
"created": "Wed, 5 Jul 2023 07:39:47 GMT"
}
] | 2023-07-06T00:00:00 | [
[
"Luo",
"Zheheng",
""
],
[
"Liu",
"Lei",
""
],
[
"Xie",
"Qianqian",
""
],
[
"Ananiadou",
"Sophia",
""
]
] | TITLE: Graph Contrastive Topic Model
ABSTRACT: Existing NTMs with contrastive learning suffer from the sample bias problem
owing to the word frequency-based sampling strategy, which may result in false
negative samples with similar semantics to the prototypes. In this paper, we
aim to explore the efficient sampling strategy and contrastive learning in NTMs
to address the aforementioned issue. We propose a new sampling assumption that
negative samples should contain words that are semantically irrelevant to the
prototype. Based on it, we propose the graph contrastive topic model (GCTM),
which conducts graph contrastive learning (GCL) using informative positive and
negative samples that are generated by the graph-based sampling strategy
leveraging in-depth correlation and irrelevance among documents and words. In
GCTM, we first model the input document as the document word bipartite graph
(DWBG), and construct positive and negative word co-occurrence graphs (WCGs),
encoded by graph neural networks, to express in-depth semantic correlation and
irrelevance among words. Based on the DWBG and WCGs, we design the
document-word information propagation (DWIP) process to perform the edge
perturbation of DWBG, based on multi-hop correlations/irrelevance among
documents and words. This yields the desired negative and positive samples,
which will be utilized for GCL together with the prototypes to improve learning
document topic representations and latent topics. We further show that GCL can
be interpreted as the structured variational graph auto-encoder which maximizes
the mutual information of latent topic representations of different
perspectives on DWBG. Experiments on several benchmark datasets demonstrate the
effectiveness of our method for topic coherence and document representation
learning compared with existing SOTA methods.
| no_new_dataset | 0.711074 |
2307.02090 | Mohan Zhou | Mohan Zhou, Yalong Bai, Wei Zhang, Ting Yao, Tiejun Zhao | Interactive Conversational Head Generation | arXiv admin note: text overlap with arXiv:2112.13548 | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We introduce a new conversation head generation benchmark for synthesizing
behaviors of a single interlocutor in a face-to-face conversation. The
capability to automatically synthesize interlocutors which can participate in
long and multi-turn conversations is vital and offer benefits for various
applications, including digital humans, virtual agents, and social robots.
While existing research primarily focuses on talking head generation (one-way
interaction), hindering the ability to create a digital human for conversation
(two-way) interaction due to the absence of listening and interaction parts. In
this work, we construct two datasets to address this issue, ``ViCo'' for
independent talking and listening head generation tasks at the sentence level,
and ``ViCo-X'', for synthesizing interlocutors in multi-turn conversational
scenarios. Based on ViCo and ViCo-X, we define three novel tasks targeting the
interaction modeling during the face-to-face conversation: 1) responsive
listening head generation making listeners respond actively to the speaker with
non-verbal signals, 2) expressive talking head generation guiding speakers to
be aware of listeners' behaviors, and 3) conversational head generation to
integrate the talking/listening ability in one interlocutor. Along with the
datasets, we also propose corresponding baseline solutions to the three
aforementioned tasks. Experimental results show that our baseline method could
generate responsive and vivid agents that can collaborate with real person to
fulfil the whole conversation. Project page: https://vico.solutions/.
| [
{
"version": "v1",
"created": "Wed, 5 Jul 2023 08:06:26 GMT"
}
] | 2023-07-06T00:00:00 | [
[
"Zhou",
"Mohan",
""
],
[
"Bai",
"Yalong",
""
],
[
"Zhang",
"Wei",
""
],
[
"Yao",
"Ting",
""
],
[
"Zhao",
"Tiejun",
""
]
] | TITLE: Interactive Conversational Head Generation
ABSTRACT: We introduce a new conversation head generation benchmark for synthesizing
behaviors of a single interlocutor in a face-to-face conversation. The
capability to automatically synthesize interlocutors which can participate in
long and multi-turn conversations is vital and offer benefits for various
applications, including digital humans, virtual agents, and social robots.
While existing research primarily focuses on talking head generation (one-way
interaction), hindering the ability to create a digital human for conversation
(two-way) interaction due to the absence of listening and interaction parts. In
this work, we construct two datasets to address this issue, ``ViCo'' for
independent talking and listening head generation tasks at the sentence level,
and ``ViCo-X'', for synthesizing interlocutors in multi-turn conversational
scenarios. Based on ViCo and ViCo-X, we define three novel tasks targeting the
interaction modeling during the face-to-face conversation: 1) responsive
listening head generation making listeners respond actively to the speaker with
non-verbal signals, 2) expressive talking head generation guiding speakers to
be aware of listeners' behaviors, and 3) conversational head generation to
integrate the talking/listening ability in one interlocutor. Along with the
datasets, we also propose corresponding baseline solutions to the three
aforementioned tasks. Experimental results show that our baseline method could
generate responsive and vivid agents that can collaborate with real person to
fulfil the whole conversation. Project page: https://vico.solutions/.
| new_dataset | 0.71881 |
2307.02094 | Adam Ivankay | Adam Ivankay, Mattia Rigotti, Pascal Frossard | DARE: Towards Robust Text Explanations in Biomedical and Healthcare
Applications | Accepted at 61st Annual Meeting of the Association for Computational
Linguistics (ACL) 2023 | null | null | null | cs.LG cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Along with the successful deployment of deep neural networks in several
application domains, the need to unravel the black-box nature of these networks
has seen a significant increase recently. Several methods have been introduced
to provide insight into the inference process of deep neural networks. However,
most of these explainability methods have been shown to be brittle in the face
of adversarial perturbations of their inputs in the image and generic textual
domain. In this work we show that this phenomenon extends to specific and
important high stakes domains like biomedical datasets. In particular, we
observe that the robustness of explanations should be characterized in terms of
the accuracy of the explanation in linking a model's inputs and its decisions -
faithfulness - and its relevance from the perspective of domain experts -
plausibility. This is crucial to prevent explanations that are inaccurate but
still look convincing in the context of the domain at hand. To this end, we
show how to adapt current attribution robustness estimation methods to a given
domain, so as to take into account domain-specific plausibility. This results
in our DomainAdaptiveAREstimator (DARE) attribution robustness estimator,
allowing us to properly characterize the domain-specific robustness of faithful
explanations. Next, we provide two methods, adversarial training and FAR
training, to mitigate the brittleness characterized by DARE, allowing us to
train networks that display robust attributions. Finally, we empirically
validate our methods with extensive experiments on three established biomedical
benchmarks.
| [
{
"version": "v1",
"created": "Wed, 5 Jul 2023 08:11:40 GMT"
}
] | 2023-07-06T00:00:00 | [
[
"Ivankay",
"Adam",
""
],
[
"Rigotti",
"Mattia",
""
],
[
"Frossard",
"Pascal",
""
]
] | TITLE: DARE: Towards Robust Text Explanations in Biomedical and Healthcare
Applications
ABSTRACT: Along with the successful deployment of deep neural networks in several
application domains, the need to unravel the black-box nature of these networks
has seen a significant increase recently. Several methods have been introduced
to provide insight into the inference process of deep neural networks. However,
most of these explainability methods have been shown to be brittle in the face
of adversarial perturbations of their inputs in the image and generic textual
domain. In this work we show that this phenomenon extends to specific and
important high stakes domains like biomedical datasets. In particular, we
observe that the robustness of explanations should be characterized in terms of
the accuracy of the explanation in linking a model's inputs and its decisions -
faithfulness - and its relevance from the perspective of domain experts -
plausibility. This is crucial to prevent explanations that are inaccurate but
still look convincing in the context of the domain at hand. To this end, we
show how to adapt current attribution robustness estimation methods to a given
domain, so as to take into account domain-specific plausibility. This results
in our DomainAdaptiveAREstimator (DARE) attribution robustness estimator,
allowing us to properly characterize the domain-specific robustness of faithful
explanations. Next, we provide two methods, adversarial training and FAR
training, to mitigate the brittleness characterized by DARE, allowing us to
train networks that display robust attributions. Finally, we empirically
validate our methods with extensive experiments on three established biomedical
benchmarks.
| no_new_dataset | 0.708792 |
2307.02120 | Kim Cheng Sheang | Kim Cheng Sheang and Horacio Saggion | Multilingual Controllable Transformer-Based Lexical Simplification | The paper is accepted for SEPLN 2023 | null | null | null | cs.CL | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Text is by far the most ubiquitous source of knowledge and information and
should be made easily accessible to as many people as possible; however, texts
often contain complex words that hinder reading comprehension and
accessibility. Therefore, suggesting simpler alternatives for complex words
without compromising meaning would help convey the information to a broader
audience. This paper proposes mTLS, a multilingual controllable
Transformer-based Lexical Simplification (LS) system fined-tuned with the T5
model. The novelty of this work lies in the use of language-specific prefixes,
control tokens, and candidates extracted from pre-trained masked language
models to learn simpler alternatives for complex words. The evaluation results
on three well-known LS datasets -- LexMTurk, BenchLS, and NNSEval -- show that
our model outperforms the previous state-of-the-art models like LSBert and
ConLS. Moreover, further evaluation of our approach on the part of the recent
TSAR-2022 multilingual LS shared-task dataset shows that our model performs
competitively when compared with the participating systems for English LS and
even outperforms the GPT-3 model on several metrics. Moreover, our model
obtains performance gains also for Spanish and Portuguese.
| [
{
"version": "v1",
"created": "Wed, 5 Jul 2023 08:48:19 GMT"
}
] | 2023-07-06T00:00:00 | [
[
"Sheang",
"Kim Cheng",
""
],
[
"Saggion",
"Horacio",
""
]
] | TITLE: Multilingual Controllable Transformer-Based Lexical Simplification
ABSTRACT: Text is by far the most ubiquitous source of knowledge and information and
should be made easily accessible to as many people as possible; however, texts
often contain complex words that hinder reading comprehension and
accessibility. Therefore, suggesting simpler alternatives for complex words
without compromising meaning would help convey the information to a broader
audience. This paper proposes mTLS, a multilingual controllable
Transformer-based Lexical Simplification (LS) system fined-tuned with the T5
model. The novelty of this work lies in the use of language-specific prefixes,
control tokens, and candidates extracted from pre-trained masked language
models to learn simpler alternatives for complex words. The evaluation results
on three well-known LS datasets -- LexMTurk, BenchLS, and NNSEval -- show that
our model outperforms the previous state-of-the-art models like LSBert and
ConLS. Moreover, further evaluation of our approach on the part of the recent
TSAR-2022 multilingual LS shared-task dataset shows that our model performs
competitively when compared with the participating systems for English LS and
even outperforms the GPT-3 model on several metrics. Moreover, our model
obtains performance gains also for Spanish and Portuguese.
| no_new_dataset | 0.703537 |
2307.02126 | Shiyu Liu | Shaogao Lv, Gang Wen, Shiyu Liu, Linsen Wei and Ming Li | Robust Graph Structure Learning with the Alignment of Features and
Adjacency Matrix | null | null | null | null | cs.LG stat.ML | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | To improve the robustness of graph neural networks (GNN), graph structure
learning (GSL) has attracted great interest due to the pervasiveness of noise
in graph data. Many approaches have been proposed for GSL to jointly learn a
clean graph structure and corresponding representations. To extend the previous
work, this paper proposes a novel regularized GSL approach, particularly with
an alignment of feature information and graph information, which is motivated
mainly by our derived lower bound of node-level Rademacher complexity for GNNs.
Additionally, our proposed approach incorporates sparse dimensional reduction
to leverage low-dimensional node features that are relevant to the graph
structure. To evaluate the effectiveness of our approach, we conduct
experiments on real-world graphs. The results demonstrate that our proposed GSL
method outperforms several competitive baselines, especially in scenarios where
the graph structures are heavily affected by noise. Overall, our research
highlights the importance of integrating feature and graph information
alignment in GSL, as inspired by our derived theoretical result, and showcases
the superiority of our approach in handling noisy graph structures through
comprehensive experiments on real-world datasets.
| [
{
"version": "v1",
"created": "Wed, 5 Jul 2023 09:05:14 GMT"
}
] | 2023-07-06T00:00:00 | [
[
"Lv",
"Shaogao",
""
],
[
"Wen",
"Gang",
""
],
[
"Liu",
"Shiyu",
""
],
[
"Wei",
"Linsen",
""
],
[
"Li",
"Ming",
""
]
] | TITLE: Robust Graph Structure Learning with the Alignment of Features and
Adjacency Matrix
ABSTRACT: To improve the robustness of graph neural networks (GNN), graph structure
learning (GSL) has attracted great interest due to the pervasiveness of noise
in graph data. Many approaches have been proposed for GSL to jointly learn a
clean graph structure and corresponding representations. To extend the previous
work, this paper proposes a novel regularized GSL approach, particularly with
an alignment of feature information and graph information, which is motivated
mainly by our derived lower bound of node-level Rademacher complexity for GNNs.
Additionally, our proposed approach incorporates sparse dimensional reduction
to leverage low-dimensional node features that are relevant to the graph
structure. To evaluate the effectiveness of our approach, we conduct
experiments on real-world graphs. The results demonstrate that our proposed GSL
method outperforms several competitive baselines, especially in scenarios where
the graph structures are heavily affected by noise. Overall, our research
highlights the importance of integrating feature and graph information
alignment in GSL, as inspired by our derived theoretical result, and showcases
the superiority of our approach in handling noisy graph structures through
comprehensive experiments on real-world datasets.
| no_new_dataset | 0.709573 |
2307.02138 | Rui Gong | Rui Gong, Martin Danelljan, Han Sun, Julio Delgado Mangas, Luc Van
Gool | Prompting Diffusion Representations for Cross-Domain Semantic
Segmentation | 17 pages, 3 figures, 11 tables | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | While originally designed for image generation, diffusion models have
recently shown to provide excellent pretrained feature representations for
semantic segmentation. Intrigued by this result, we set out to explore how well
diffusion-pretrained representations generalize to new domains, a crucial
ability for any representation. We find that diffusion-pretraining achieves
extraordinary domain generalization results for semantic segmentation,
outperforming both supervised and self-supervised backbone networks. Motivated
by this, we investigate how to utilize the model's unique ability of taking an
input prompt, in order to further enhance its cross-domain performance. We
introduce a scene prompt and a prompt randomization strategy to help further
disentangle the domain-invariant information when training the segmentation
head. Moreover, we propose a simple but highly effective approach for test-time
domain adaptation, based on learning a scene prompt on the target domain in an
unsupervised manner. Extensive experiments conducted on four synthetic-to-real
and clear-to-adverse weather benchmarks demonstrate the effectiveness of our
approaches. Without resorting to any complex techniques, such as image
translation, augmentation, or rare-class sampling, we set a new
state-of-the-art on all benchmarks. Our implementation will be publicly
available at \url{https://github.com/ETHRuiGong/PTDiffSeg}.
| [
{
"version": "v1",
"created": "Wed, 5 Jul 2023 09:28:25 GMT"
}
] | 2023-07-06T00:00:00 | [
[
"Gong",
"Rui",
""
],
[
"Danelljan",
"Martin",
""
],
[
"Sun",
"Han",
""
],
[
"Mangas",
"Julio Delgado",
""
],
[
"Van Gool",
"Luc",
""
]
] | TITLE: Prompting Diffusion Representations for Cross-Domain Semantic
Segmentation
ABSTRACT: While originally designed for image generation, diffusion models have
recently shown to provide excellent pretrained feature representations for
semantic segmentation. Intrigued by this result, we set out to explore how well
diffusion-pretrained representations generalize to new domains, a crucial
ability for any representation. We find that diffusion-pretraining achieves
extraordinary domain generalization results for semantic segmentation,
outperforming both supervised and self-supervised backbone networks. Motivated
by this, we investigate how to utilize the model's unique ability of taking an
input prompt, in order to further enhance its cross-domain performance. We
introduce a scene prompt and a prompt randomization strategy to help further
disentangle the domain-invariant information when training the segmentation
head. Moreover, we propose a simple but highly effective approach for test-time
domain adaptation, based on learning a scene prompt on the target domain in an
unsupervised manner. Extensive experiments conducted on four synthetic-to-real
and clear-to-adverse weather benchmarks demonstrate the effectiveness of our
approaches. Without resorting to any complex techniques, such as image
translation, augmentation, or rare-class sampling, we set a new
state-of-the-art on all benchmarks. Our implementation will be publicly
available at \url{https://github.com/ETHRuiGong/PTDiffSeg}.
| no_new_dataset | 0.709849 |
2307.02157 | Zhi Zheng | Zhi Zheng, Zhaopeng Qiu, Xiao Hu, Likang Wu, Hengshu Zhu, Hui Xiong | Generative Job Recommendations with Large Language Model | null | null | null | null | cs.IR cs.CL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The rapid development of online recruitment services has encouraged the
utilization of recommender systems to streamline the job seeking process.
Predominantly, current job recommendations deploy either collaborative
filtering or person-job matching strategies. However, these models tend to
operate as "black-box" systems and lack the capacity to offer explainable
guidance to job seekers. Moreover, conventional matching-based recommendation
methods are limited to retrieving and ranking existing jobs in the database,
restricting their potential as comprehensive career AI advisors. To this end,
here we present GIRL (GeneratIve job Recommendation based on Large language
models), a novel approach inspired by recent advancements in the field of Large
Language Models (LLMs). We initially employ a Supervised Fine-Tuning (SFT)
strategy to instruct the LLM-based generator in crafting suitable Job
Descriptions (JDs) based on the Curriculum Vitae (CV) of a job seeker.
Moreover, we propose to train a model which can evaluate the matching degree
between CVs and JDs as a reward model, and we use Proximal Policy Optimization
(PPO)-based Reinforcement Learning (RL) method to further fine-tine the
generator. This aligns the generator with recruiter feedback, tailoring the
output to better meet employer preferences. In particular, GIRL serves as a job
seeker-centric generative model, providing job suggestions without the need of
a candidate set. This capability also enhances the performance of existing job
recommendation models by supplementing job seeking features with generated
content. With extensive experiments on a large-scale real-world dataset, we
demonstrate the substantial effectiveness of our approach. We believe that GIRL
introduces a paradigm-shifting approach to job recommendation systems,
fostering a more personalized and comprehensive job-seeking experience.
| [
{
"version": "v1",
"created": "Wed, 5 Jul 2023 09:58:08 GMT"
}
] | 2023-07-06T00:00:00 | [
[
"Zheng",
"Zhi",
""
],
[
"Qiu",
"Zhaopeng",
""
],
[
"Hu",
"Xiao",
""
],
[
"Wu",
"Likang",
""
],
[
"Zhu",
"Hengshu",
""
],
[
"Xiong",
"Hui",
""
]
] | TITLE: Generative Job Recommendations with Large Language Model
ABSTRACT: The rapid development of online recruitment services has encouraged the
utilization of recommender systems to streamline the job seeking process.
Predominantly, current job recommendations deploy either collaborative
filtering or person-job matching strategies. However, these models tend to
operate as "black-box" systems and lack the capacity to offer explainable
guidance to job seekers. Moreover, conventional matching-based recommendation
methods are limited to retrieving and ranking existing jobs in the database,
restricting their potential as comprehensive career AI advisors. To this end,
here we present GIRL (GeneratIve job Recommendation based on Large language
models), a novel approach inspired by recent advancements in the field of Large
Language Models (LLMs). We initially employ a Supervised Fine-Tuning (SFT)
strategy to instruct the LLM-based generator in crafting suitable Job
Descriptions (JDs) based on the Curriculum Vitae (CV) of a job seeker.
Moreover, we propose to train a model which can evaluate the matching degree
between CVs and JDs as a reward model, and we use Proximal Policy Optimization
(PPO)-based Reinforcement Learning (RL) method to further fine-tine the
generator. This aligns the generator with recruiter feedback, tailoring the
output to better meet employer preferences. In particular, GIRL serves as a job
seeker-centric generative model, providing job suggestions without the need of
a candidate set. This capability also enhances the performance of existing job
recommendation models by supplementing job seeking features with generated
content. With extensive experiments on a large-scale real-world dataset, we
demonstrate the substantial effectiveness of our approach. We believe that GIRL
introduces a paradigm-shifting approach to job recommendation systems,
fostering a more personalized and comprehensive job-seeking experience.
| no_new_dataset | 0.708395 |
2307.02200 | Liu Shanqi | Shanqi Liu, Weiwei Liu, Wenzhou Chen, Guanzhong Tian, Yong Liu | Multi-Agent Cooperation via Unsupervised Learning of Joint Intentions | null | null | null | null | cs.MA | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The field of cooperative multi-agent reinforcement learning (MARL) has seen
widespread use in addressing complex coordination tasks. While value
decomposition methods in MARL have been popular, they have limitations in
solving tasks with non-monotonic returns, restricting their general
application. Our work highlights the significance of joint intentions in
cooperation, which can overcome non-monotonic problems and increase the
interpretability of the learning process. To this end, we present a novel MARL
method that leverages learnable joint intentions. Our method employs a
hierarchical framework consisting of a joint intention policy and a behavior
policy to formulate the optimal cooperative policy. The joint intentions are
autonomously learned in a latent space through unsupervised learning and enable
the method adaptable to different agent configurations. Our results demonstrate
significant performance improvements in both the StarCraft micromanagement
benchmark and challenging MAgent domains, showcasing the effectiveness of our
method in learning meaningful joint intentions.
| [
{
"version": "v1",
"created": "Wed, 5 Jul 2023 10:52:02 GMT"
}
] | 2023-07-06T00:00:00 | [
[
"Liu",
"Shanqi",
""
],
[
"Liu",
"Weiwei",
""
],
[
"Chen",
"Wenzhou",
""
],
[
"Tian",
"Guanzhong",
""
],
[
"Liu",
"Yong",
""
]
] | TITLE: Multi-Agent Cooperation via Unsupervised Learning of Joint Intentions
ABSTRACT: The field of cooperative multi-agent reinforcement learning (MARL) has seen
widespread use in addressing complex coordination tasks. While value
decomposition methods in MARL have been popular, they have limitations in
solving tasks with non-monotonic returns, restricting their general
application. Our work highlights the significance of joint intentions in
cooperation, which can overcome non-monotonic problems and increase the
interpretability of the learning process. To this end, we present a novel MARL
method that leverages learnable joint intentions. Our method employs a
hierarchical framework consisting of a joint intention policy and a behavior
policy to formulate the optimal cooperative policy. The joint intentions are
autonomously learned in a latent space through unsupervised learning and enable
the method adaptable to different agent configurations. Our results demonstrate
significant performance improvements in both the StarCraft micromanagement
benchmark and challenging MAgent domains, showcasing the effectiveness of our
method in learning meaningful joint intentions.
| no_new_dataset | 0.709648 |
2307.02238 | Subhradeep Kayal | Shuai Chen and Subhradeep Kayal and Marleen de Bruijne | Source Identification: A Self-Supervision Task for Dense Prediction | Under review | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | The paradigm of self-supervision focuses on representation learning from raw
data without the need of labor-consuming annotations, which is the main
bottleneck of current data-driven methods. Self-supervision tasks are often
used to pre-train a neural network with a large amount of unlabeled data and
extract generic features of the dataset. The learned model is likely to contain
useful information which can be transferred to the downstream main task and
improve performance compared to random parameter initialization. In this paper,
we propose a new self-supervision task called source identification (SI), which
is inspired by the classic blind source separation problem. Synthetic images
are generated by fusing multiple source images and the network's task is to
reconstruct the original images, given the fused images. A proper understanding
of the image content is required to successfully solve the task. We validate
our method on two medical image segmentation tasks: brain tumor segmentation
and white matter hyperintensities segmentation. The results show that the
proposed SI task outperforms traditional self-supervision tasks for dense
predictions including inpainting, pixel shuffling, intensity shift, and
super-resolution. Among variations of the SI task fusing images of different
types, fusing images from different patients performs best.
| [
{
"version": "v1",
"created": "Wed, 5 Jul 2023 12:27:58 GMT"
}
] | 2023-07-06T00:00:00 | [
[
"Chen",
"Shuai",
""
],
[
"Kayal",
"Subhradeep",
""
],
[
"de Bruijne",
"Marleen",
""
]
] | TITLE: Source Identification: A Self-Supervision Task for Dense Prediction
ABSTRACT: The paradigm of self-supervision focuses on representation learning from raw
data without the need of labor-consuming annotations, which is the main
bottleneck of current data-driven methods. Self-supervision tasks are often
used to pre-train a neural network with a large amount of unlabeled data and
extract generic features of the dataset. The learned model is likely to contain
useful information which can be transferred to the downstream main task and
improve performance compared to random parameter initialization. In this paper,
we propose a new self-supervision task called source identification (SI), which
is inspired by the classic blind source separation problem. Synthetic images
are generated by fusing multiple source images and the network's task is to
reconstruct the original images, given the fused images. A proper understanding
of the image content is required to successfully solve the task. We validate
our method on two medical image segmentation tasks: brain tumor segmentation
and white matter hyperintensities segmentation. The results show that the
proposed SI task outperforms traditional self-supervision tasks for dense
predictions including inpainting, pixel shuffling, intensity shift, and
super-resolution. Among variations of the SI task fusing images of different
types, fusing images from different patients performs best.
| no_new_dataset | 0.708036 |
2307.02244 | Sandipana Dowerah | Sandipana Dowerah, Ajinkya Kulkarni, Romain Serizel (MULTISPEECH),
Denis Jouvet | Self-supervised learning with diffusion-based multichannel speech
enhancement for speaker verification under noisy conditions | null | null | null | null | cs.SD eess.AS | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The paper introduces Diff-Filter, a multichannel speech enhancement approach
based on the diffusion probabilistic model, for improving speaker verification
performance under noisy and reverberant conditions. It also presents a new
two-step training procedure that takes the benefit of self-supervised learning.
In the first stage, the Diff-Filter is trained by conducting timedomain speech
filtering using a scoring-based diffusion model. In the second stage, the
Diff-Filter is jointly optimized with a pre-trained ECAPA-TDNN speaker
verification model under a self-supervised learning framework. We present a
novel loss based on equal error rate. This loss is used to conduct
selfsupervised learning on a dataset that is not labelled in terms of speakers.
The proposed approach is evaluated on MultiSV, a multichannel speaker
verification dataset, and shows significant improvements in performance under
noisy multichannel conditions.
| [
{
"version": "v1",
"created": "Wed, 5 Jul 2023 12:36:39 GMT"
}
] | 2023-07-06T00:00:00 | [
[
"Dowerah",
"Sandipana",
"",
"MULTISPEECH"
],
[
"Kulkarni",
"Ajinkya",
"",
"MULTISPEECH"
],
[
"Serizel",
"Romain",
"",
"MULTISPEECH"
],
[
"Jouvet",
"Denis",
""
]
] | TITLE: Self-supervised learning with diffusion-based multichannel speech
enhancement for speaker verification under noisy conditions
ABSTRACT: The paper introduces Diff-Filter, a multichannel speech enhancement approach
based on the diffusion probabilistic model, for improving speaker verification
performance under noisy and reverberant conditions. It also presents a new
two-step training procedure that takes the benefit of self-supervised learning.
In the first stage, the Diff-Filter is trained by conducting timedomain speech
filtering using a scoring-based diffusion model. In the second stage, the
Diff-Filter is jointly optimized with a pre-trained ECAPA-TDNN speaker
verification model under a self-supervised learning framework. We present a
novel loss based on equal error rate. This loss is used to conduct
selfsupervised learning on a dataset that is not labelled in terms of speakers.
The proposed approach is evaluated on MultiSV, a multichannel speaker
verification dataset, and shows significant improvements in performance under
noisy multichannel conditions.
| no_new_dataset | 0.694704 |
2307.02246 | Jayateja Kalla | Jayateja Kalla and Soma Biswas | S3C: Self-Supervised Stochastic Classifiers for Few-Shot
Class-Incremental Learning | Accepted in ECCV 2022 | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | Few-shot class-incremental learning (FSCIL) aims to learn progressively about
new classes with very few labeled samples, without forgetting the knowledge of
already learnt classes. FSCIL suffers from two major challenges: (i)
over-fitting on the new classes due to limited amount of data, (ii)
catastrophically forgetting about the old classes due to unavailability of data
from these classes in the incremental stages. In this work, we propose a
self-supervised stochastic classifier (S3C) to counter both these challenges in
FSCIL. The stochasticity of the classifier weights (or class prototypes) not
only mitigates the adverse effect of absence of large number of samples of the
new classes, but also the absence of samples from previously learnt classes
during the incremental steps. This is complemented by the self-supervision
component, which helps to learn features from the base classes which generalize
well to unseen classes that are encountered in future, thus reducing
catastrophic forgetting. Extensive evaluation on three benchmark datasets using
multiple evaluation metrics show the effectiveness of the proposed framework.
We also experiment on two additional realistic scenarios of FSCIL, namely where
the number of annotated data available for each of the new classes can be
different, and also where the number of base classes is much lesser, and show
that the proposed S3C performs significantly better than the state-of-the-art
for all these challenging scenarios.
| [
{
"version": "v1",
"created": "Wed, 5 Jul 2023 12:41:46 GMT"
}
] | 2023-07-06T00:00:00 | [
[
"Kalla",
"Jayateja",
""
],
[
"Biswas",
"Soma",
""
]
] | TITLE: S3C: Self-Supervised Stochastic Classifiers for Few-Shot
Class-Incremental Learning
ABSTRACT: Few-shot class-incremental learning (FSCIL) aims to learn progressively about
new classes with very few labeled samples, without forgetting the knowledge of
already learnt classes. FSCIL suffers from two major challenges: (i)
over-fitting on the new classes due to limited amount of data, (ii)
catastrophically forgetting about the old classes due to unavailability of data
from these classes in the incremental stages. In this work, we propose a
self-supervised stochastic classifier (S3C) to counter both these challenges in
FSCIL. The stochasticity of the classifier weights (or class prototypes) not
only mitigates the adverse effect of absence of large number of samples of the
new classes, but also the absence of samples from previously learnt classes
during the incremental steps. This is complemented by the self-supervision
component, which helps to learn features from the base classes which generalize
well to unseen classes that are encountered in future, thus reducing
catastrophic forgetting. Extensive evaluation on three benchmark datasets using
multiple evaluation metrics show the effectiveness of the proposed framework.
We also experiment on two additional realistic scenarios of FSCIL, namely where
the number of annotated data available for each of the new classes can be
different, and also where the number of base classes is much lesser, and show
that the proposed S3C performs significantly better than the state-of-the-art
for all these challenging scenarios.
| no_new_dataset | 0.71262 |
2307.02269 | Lasha Abzianidze | Lasha Abzianidze, Joost Zwarts, Yoad Winter | SpaceNLI: Evaluating the Consistency of Predicting Inferences in Space | Accepted and presented at the NALOMA (Natural Logic Meets Machine
Learning) workshop. The paper repository is at
https://github.com/kovvalsky/SpaceNLI | null | null | null | cs.CL | http://creativecommons.org/licenses/by/4.0/ | While many natural language inference (NLI) datasets target certain semantic
phenomena, e.g., negation, tense & aspect, monotonicity, and presupposition, to
the best of our knowledge, there is no NLI dataset that involves diverse types
of spatial expressions and reasoning. We fill this gap by semi-automatically
creating an NLI dataset for spatial reasoning, called SpaceNLI. The data
samples are automatically generated from a curated set of reasoning patterns,
where the patterns are annotated with inference labels by experts. We test
several SOTA NLI systems on SpaceNLI to gauge the complexity of the dataset and
the system's capacity for spatial reasoning. Moreover, we introduce a Pattern
Accuracy and argue that it is a more reliable and stricter measure than the
accuracy for evaluating a system's performance on pattern-based generated data
samples. Based on the evaluation results we find that the systems obtain
moderate results on the spatial NLI problems but lack consistency per inference
pattern. The results also reveal that non-projective spatial inferences
(especially due to the "between" preposition) are the most challenging ones.
| [
{
"version": "v1",
"created": "Wed, 5 Jul 2023 13:08:18 GMT"
}
] | 2023-07-06T00:00:00 | [
[
"Abzianidze",
"Lasha",
""
],
[
"Zwarts",
"Joost",
""
],
[
"Winter",
"Yoad",
""
]
] | TITLE: SpaceNLI: Evaluating the Consistency of Predicting Inferences in Space
ABSTRACT: While many natural language inference (NLI) datasets target certain semantic
phenomena, e.g., negation, tense & aspect, monotonicity, and presupposition, to
the best of our knowledge, there is no NLI dataset that involves diverse types
of spatial expressions and reasoning. We fill this gap by semi-automatically
creating an NLI dataset for spatial reasoning, called SpaceNLI. The data
samples are automatically generated from a curated set of reasoning patterns,
where the patterns are annotated with inference labels by experts. We test
several SOTA NLI systems on SpaceNLI to gauge the complexity of the dataset and
the system's capacity for spatial reasoning. Moreover, we introduce a Pattern
Accuracy and argue that it is a more reliable and stricter measure than the
accuracy for evaluating a system's performance on pattern-based generated data
samples. Based on the evaluation results we find that the systems obtain
moderate results on the spatial NLI problems but lack consistency per inference
pattern. The results also reveal that non-projective spatial inferences
(especially due to the "between" preposition) are the most challenging ones.
| new_dataset | 0.71484 |
2307.02270 | Yuguang Shi | Yuguang Shi | SVDM: Single-View Diffusion Model for Pseudo-Stereo 3D Object Detection | arXiv admin note: text overlap with arXiv:2203.02112,
arXiv:2303.01469 by other authors | null | null | null | cs.CV cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | One of the key problems in 3D object detection is to reduce the accuracy gap
between methods based on LiDAR sensors and those based on monocular cameras. A
recently proposed framework for monocular 3D detection based on Pseudo-Stereo
has received considerable attention in the community. However, so far these two
problems are discovered in existing practices, including (1) monocular depth
estimation and Pseudo-Stereo detector must be trained separately, (2) Difficult
to be compatible with different stereo detectors and (3) the overall
calculation is large, which affects the reasoning speed. In this work, we
propose an end-to-end, efficient pseudo-stereo 3D detection framework by
introducing a Single-View Diffusion Model (SVDM) that uses a few iterations to
gradually deliver right informative pixels to the left image. SVDM allows the
entire pseudo-stereo 3D detection pipeline to be trained end-to-end and can
benefit from the training of stereo detectors. Afterwards, we further explore
the application of SVDM in depth-free stereo 3D detection, and the final
framework is compatible with most stereo detectors. Among multiple benchmarks
on the KITTI dataset, we achieve new state-of-the-art performance.
| [
{
"version": "v1",
"created": "Wed, 5 Jul 2023 13:10:37 GMT"
}
] | 2023-07-06T00:00:00 | [
[
"Shi",
"Yuguang",
""
]
] | TITLE: SVDM: Single-View Diffusion Model for Pseudo-Stereo 3D Object Detection
ABSTRACT: One of the key problems in 3D object detection is to reduce the accuracy gap
between methods based on LiDAR sensors and those based on monocular cameras. A
recently proposed framework for monocular 3D detection based on Pseudo-Stereo
has received considerable attention in the community. However, so far these two
problems are discovered in existing practices, including (1) monocular depth
estimation and Pseudo-Stereo detector must be trained separately, (2) Difficult
to be compatible with different stereo detectors and (3) the overall
calculation is large, which affects the reasoning speed. In this work, we
propose an end-to-end, efficient pseudo-stereo 3D detection framework by
introducing a Single-View Diffusion Model (SVDM) that uses a few iterations to
gradually deliver right informative pixels to the left image. SVDM allows the
entire pseudo-stereo 3D detection pipeline to be trained end-to-end and can
benefit from the training of stereo detectors. Afterwards, we further explore
the application of SVDM in depth-free stereo 3D detection, and the final
framework is compatible with most stereo detectors. Among multiple benchmarks
on the KITTI dataset, we achieve new state-of-the-art performance.
| no_new_dataset | 0.709422 |
2307.02280 | Kun Li | Kun Li, George Vosselman, Michael Ying Yang | Interactive Image Segmentation with Cross-Modality Vision Transformers | 16 pages | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Interactive image segmentation aims to segment the target from the background
with the manual guidance, which takes as input multimodal data such as images,
clicks, scribbles, and bounding boxes. Recently, vision transformers have
achieved a great success in several downstream visual tasks, and a few efforts
have been made to bring this powerful architecture to interactive segmentation
task. However, the previous works neglect the relations between two modalities
and directly mock the way of processing purely visual information with
self-attentions. In this paper, we propose a simple yet effective network for
click-based interactive segmentation with cross-modality vision transformers.
Cross-modality transformers exploits mutual information to better guide the
learning process. The experiments on several benchmarks show that the proposed
method achieves superior performance in comparison to the previous
state-of-the-art models. The stability of our method in term of avoiding
failure cases shows its potential to be a practical annotation tool. The code
and pretrained models will be released under
https://github.com/lik1996/iCMFormer.
| [
{
"version": "v1",
"created": "Wed, 5 Jul 2023 13:29:05 GMT"
}
] | 2023-07-06T00:00:00 | [
[
"Li",
"Kun",
""
],
[
"Vosselman",
"George",
""
],
[
"Yang",
"Michael Ying",
""
]
] | TITLE: Interactive Image Segmentation with Cross-Modality Vision Transformers
ABSTRACT: Interactive image segmentation aims to segment the target from the background
with the manual guidance, which takes as input multimodal data such as images,
clicks, scribbles, and bounding boxes. Recently, vision transformers have
achieved a great success in several downstream visual tasks, and a few efforts
have been made to bring this powerful architecture to interactive segmentation
task. However, the previous works neglect the relations between two modalities
and directly mock the way of processing purely visual information with
self-attentions. In this paper, we propose a simple yet effective network for
click-based interactive segmentation with cross-modality vision transformers.
Cross-modality transformers exploits mutual information to better guide the
learning process. The experiments on several benchmarks show that the proposed
method achieves superior performance in comparison to the previous
state-of-the-art models. The stability of our method in term of avoiding
failure cases shows its potential to be a practical annotation tool. The code
and pretrained models will be released under
https://github.com/lik1996/iCMFormer.
| no_new_dataset | 0.710409 |
2307.02289 | Tai D. Nguyen | Tai D. Nguyen, Long H. Pham, Jun Sun | Fuzzing with Quantitative and Adaptive Hot-Bytes Identification | null | null | null | null | cs.CR cs.SE | http://creativecommons.org/licenses/by/4.0/ | Fuzzing has emerged as a powerful technique for finding security bugs in
complicated real-world applications. American fuzzy lop (AFL), a leading
fuzzing tool, has demonstrated its powerful bug finding ability through a vast
number of reported CVEs. However, its random mutation strategy is unable to
generate test inputs that satisfy complicated branching conditions (e.g.,
magic-byte comparisons, checksum tests, and nested if-statements), which are
commonly used in image decoders/encoders, XML parsers, and checksum tools.
Existing approaches (such as Steelix and Neuzz) on addressing this problem
assume unrealistic assumptions such as we can satisfy the branch condition
byte-to-byte or we can identify and focus on the important bytes in the input
(called hot-bytes) once and for all. In this work, we propose an approach
called \tool~which is designed based on the following principles. First, there
is a complicated relation between inputs and branching conditions and thus we
need not only an expressive model to capture such relationship but also an
informative measure so that we can learn such relationship effectively. Second,
different branching conditions demand different hot-bytes and we must adjust
our fuzzing strategy adaptively depending on which branches are the current
bottleneck. We implement our approach as an open source project and compare its
efficiency with other state-of-the-art fuzzers. Our evaluation results on 10
real-world programs and LAVA-M dataset show that \tool~achieves sustained
increases in branch coverage and discovers more bugs than other fuzzers.
| [
{
"version": "v1",
"created": "Wed, 5 Jul 2023 13:41:35 GMT"
}
] | 2023-07-06T00:00:00 | [
[
"Nguyen",
"Tai D.",
""
],
[
"Pham",
"Long H.",
""
],
[
"Sun",
"Jun",
""
]
] | TITLE: Fuzzing with Quantitative and Adaptive Hot-Bytes Identification
ABSTRACT: Fuzzing has emerged as a powerful technique for finding security bugs in
complicated real-world applications. American fuzzy lop (AFL), a leading
fuzzing tool, has demonstrated its powerful bug finding ability through a vast
number of reported CVEs. However, its random mutation strategy is unable to
generate test inputs that satisfy complicated branching conditions (e.g.,
magic-byte comparisons, checksum tests, and nested if-statements), which are
commonly used in image decoders/encoders, XML parsers, and checksum tools.
Existing approaches (such as Steelix and Neuzz) on addressing this problem
assume unrealistic assumptions such as we can satisfy the branch condition
byte-to-byte or we can identify and focus on the important bytes in the input
(called hot-bytes) once and for all. In this work, we propose an approach
called \tool~which is designed based on the following principles. First, there
is a complicated relation between inputs and branching conditions and thus we
need not only an expressive model to capture such relationship but also an
informative measure so that we can learn such relationship effectively. Second,
different branching conditions demand different hot-bytes and we must adjust
our fuzzing strategy adaptively depending on which branches are the current
bottleneck. We implement our approach as an open source project and compare its
efficiency with other state-of-the-art fuzzers. Our evaluation results on 10
real-world programs and LAVA-M dataset show that \tool~achieves sustained
increases in branch coverage and discovers more bugs than other fuzzers.
| no_new_dataset | 0.70878 |
2307.02308 | Saisai Ding | Saisai Ding, Jun Wang, Juncheng Li, and Jun Shi | Multi-Scale Prototypical Transformer for Whole Slide Image
Classification | null | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | Whole slide image (WSI) classification is an essential task in computational
pathology. Despite the recent advances in multiple instance learning (MIL) for
WSI classification, accurate classification of WSIs remains challenging due to
the extreme imbalance between the positive and negative instances in bags, and
the complicated pre-processing to fuse multi-scale information of WSI. To this
end, we propose a novel multi-scale prototypical Transformer (MSPT) for WSI
classification, which includes a prototypical Transformer (PT) module and a
multi-scale feature fusion module (MFFM). The PT is developed to reduce
redundant instances in bags by integrating prototypical learning into the
Transformer architecture. It substitutes all instances with cluster prototypes,
which are then re-calibrated through the self-attention mechanism of the
Trans-former. Thereafter, an MFFM is proposed to fuse the clustered prototypes
of different scales, which employs MLP-Mixer to enhance the information
communication between prototypes. The experimental results on two public WSI
datasets demonstrate that the proposed MSPT outperforms all the compared
algorithms, suggesting its potential applications.
| [
{
"version": "v1",
"created": "Wed, 5 Jul 2023 14:10:29 GMT"
}
] | 2023-07-06T00:00:00 | [
[
"Ding",
"Saisai",
""
],
[
"Wang",
"Jun",
""
],
[
"Li",
"Juncheng",
""
],
[
"Shi",
"Jun",
""
]
] | TITLE: Multi-Scale Prototypical Transformer for Whole Slide Image
Classification
ABSTRACT: Whole slide image (WSI) classification is an essential task in computational
pathology. Despite the recent advances in multiple instance learning (MIL) for
WSI classification, accurate classification of WSIs remains challenging due to
the extreme imbalance between the positive and negative instances in bags, and
the complicated pre-processing to fuse multi-scale information of WSI. To this
end, we propose a novel multi-scale prototypical Transformer (MSPT) for WSI
classification, which includes a prototypical Transformer (PT) module and a
multi-scale feature fusion module (MFFM). The PT is developed to reduce
redundant instances in bags by integrating prototypical learning into the
Transformer architecture. It substitutes all instances with cluster prototypes,
which are then re-calibrated through the self-attention mechanism of the
Trans-former. Thereafter, an MFFM is proposed to fuse the clustered prototypes
of different scales, which employs MLP-Mixer to enhance the information
communication between prototypes. The experimental results on two public WSI
datasets demonstrate that the proposed MSPT outperforms all the compared
algorithms, suggesting its potential applications.
| no_new_dataset | 0.707998 |
2307.02340 | Timo Pierre Schrader | Timo Pierre Schrader, Teresa B\"urkle, Sophie Henning, Sherry Tan,
Matteo Finco, Stefan Gr\"unewald, Maira Indrikova, Felix Hildebrand,
Annemarie Friedrich | MuLMS-AZ: An Argumentative Zoning Dataset for the Materials Science
Domain | 15 pages, 2 figures, 14 tables, to be published in "Proceedings of
the 4th Workshop on Computational Approaches to Discourse" | null | null | null | cs.CL | http://creativecommons.org/licenses/by-sa/4.0/ | Scientific publications follow conventionalized rhetorical structures.
Classifying the Argumentative Zone (AZ), e.g., identifying whether a sentence
states a Motivation, a Result or Background information, has been proposed to
improve processing of scholarly documents. In this work, we adapt and extend
this idea to the domain of materials science research. We present and release a
new dataset of 50 manually annotated research articles. The dataset spans seven
sub-topics and is annotated with a materials-science focused multi-label
annotation scheme for AZ. We detail corpus statistics and demonstrate high
inter-annotator agreement. Our computational experiments show that using
domain-specific pre-trained transformer-based text encoders is key to high
classification performance. We also find that AZ categories from existing
datasets in other domains are transferable to varying degrees.
| [
{
"version": "v1",
"created": "Wed, 5 Jul 2023 14:55:18 GMT"
}
] | 2023-07-06T00:00:00 | [
[
"Schrader",
"Timo Pierre",
""
],
[
"Bürkle",
"Teresa",
""
],
[
"Henning",
"Sophie",
""
],
[
"Tan",
"Sherry",
""
],
[
"Finco",
"Matteo",
""
],
[
"Grünewald",
"Stefan",
""
],
[
"Indrikova",
"Maira",
""
],
[
"Hildebrand",
"Felix",
""
],
[
"Friedrich",
"Annemarie",
""
]
] | TITLE: MuLMS-AZ: An Argumentative Zoning Dataset for the Materials Science
Domain
ABSTRACT: Scientific publications follow conventionalized rhetorical structures.
Classifying the Argumentative Zone (AZ), e.g., identifying whether a sentence
states a Motivation, a Result or Background information, has been proposed to
improve processing of scholarly documents. In this work, we adapt and extend
this idea to the domain of materials science research. We present and release a
new dataset of 50 manually annotated research articles. The dataset spans seven
sub-topics and is annotated with a materials-science focused multi-label
annotation scheme for AZ. We detail corpus statistics and demonstrate high
inter-annotator agreement. Our computational experiments show that using
domain-specific pre-trained transformer-based text encoders is key to high
classification performance. We also find that AZ categories from existing
datasets in other domains are transferable to varying degrees.
| new_dataset | 0.715772 |
2307.02367 | Steven Goldenberg | Steven Goldenberg, Malachi Schram, Kishansingh Rajput, Thomas Britton,
Chris Pappas, Dan Lu, Jared Walden, Majdi I. Radaideh, Sarah Cousineau,
Sudarshan Harave | Distance Preserving Machine Learning for Uncertainty Aware Accelerator
Capacitance Predictions | null | null | null | null | cs.LG physics.acc-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Providing accurate uncertainty estimations is essential for producing
reliable machine learning models, especially in safety-critical applications
such as accelerator systems. Gaussian process models are generally regarded as
the gold standard method for this task, but they can struggle with large,
high-dimensional datasets. Combining deep neural networks with Gaussian process
approximation techniques have shown promising results, but dimensionality
reduction through standard deep neural network layers is not guaranteed to
maintain the distance information necessary for Gaussian process models. We
build on previous work by comparing the use of the singular value decomposition
against a spectral-normalized dense layer as a feature extractor for a deep
neural Gaussian process approximation model and apply it to a capacitance
prediction problem for the High Voltage Converter Modulators in the Oak Ridge
Spallation Neutron Source. Our model shows improved distance preservation and
predicts in-distribution capacitance values with less than 1% error.
| [
{
"version": "v1",
"created": "Wed, 5 Jul 2023 15:32:39 GMT"
}
] | 2023-07-06T00:00:00 | [
[
"Goldenberg",
"Steven",
""
],
[
"Schram",
"Malachi",
""
],
[
"Rajput",
"Kishansingh",
""
],
[
"Britton",
"Thomas",
""
],
[
"Pappas",
"Chris",
""
],
[
"Lu",
"Dan",
""
],
[
"Walden",
"Jared",
""
],
[
"Radaideh",
"Majdi I.",
""
],
[
"Cousineau",
"Sarah",
""
],
[
"Harave",
"Sudarshan",
""
]
] | TITLE: Distance Preserving Machine Learning for Uncertainty Aware Accelerator
Capacitance Predictions
ABSTRACT: Providing accurate uncertainty estimations is essential for producing
reliable machine learning models, especially in safety-critical applications
such as accelerator systems. Gaussian process models are generally regarded as
the gold standard method for this task, but they can struggle with large,
high-dimensional datasets. Combining deep neural networks with Gaussian process
approximation techniques have shown promising results, but dimensionality
reduction through standard deep neural network layers is not guaranteed to
maintain the distance information necessary for Gaussian process models. We
build on previous work by comparing the use of the singular value decomposition
against a spectral-normalized dense layer as a feature extractor for a deep
neural Gaussian process approximation model and apply it to a capacitance
prediction problem for the High Voltage Converter Modulators in the Oak Ridge
Spallation Neutron Source. Our model shows improved distance preservation and
predicts in-distribution capacitance values with less than 1% error.
| no_new_dataset | 0.709831 |
2307.02379 | Alessandro Ingrosso | Alessandro Ingrosso and Emanuele Panizon | Machine learning at the mesoscale: a computation-dissipation bottleneck | 12 pages, 5 figures | null | null | null | cond-mat.stat-mech cond-mat.dis-nn cs.LG | http://creativecommons.org/licenses/by/4.0/ | The cost of information processing in physical systems calls for a trade-off
between performance and energetic expenditure. Here we formulate and study a
computation-dissipation bottleneck in mesoscopic systems used as input-output
devices. Using both real datasets and synthetic tasks, we show how
non-equilibrium leads to enhanced performance. Our framework sheds light on a
crucial compromise between information compression, input-output computation
and dynamic irreversibility induced by non-reciprocal interactions.
| [
{
"version": "v1",
"created": "Wed, 5 Jul 2023 15:46:07 GMT"
}
] | 2023-07-06T00:00:00 | [
[
"Ingrosso",
"Alessandro",
""
],
[
"Panizon",
"Emanuele",
""
]
] | TITLE: Machine learning at the mesoscale: a computation-dissipation bottleneck
ABSTRACT: The cost of information processing in physical systems calls for a trade-off
between performance and energetic expenditure. Here we formulate and study a
computation-dissipation bottleneck in mesoscopic systems used as input-output
devices. Using both real datasets and synthetic tasks, we show how
non-equilibrium leads to enhanced performance. Our framework sheds light on a
crucial compromise between information compression, input-output computation
and dynamic irreversibility induced by non-reciprocal interactions.
| no_new_dataset | 0.708855 |
2307.02392 | Renato Sortino | Renato Sortino, Thomas Cecconello, Andrea DeMarco, Giuseppe Fiameni,
Andrea Pilzer, Andrew M. Hopkins, Daniel Magro, Simone Riggi, Eva Sciacca,
Adriano Ingallinera, Cristobal Bordiu, Filomena Bufano, Concetto Spampinato | RADiff: Controllable Diffusion Models for Radio Astronomical Maps
Generation | null | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | Along with the nearing completion of the Square Kilometre Array (SKA), comes
an increasing demand for accurate and reliable automated solutions to extract
valuable information from the vast amount of data it will allow acquiring.
Automated source finding is a particularly important task in this context, as
it enables the detection and classification of astronomical objects.
Deep-learning-based object detection and semantic segmentation models have
proven to be suitable for this purpose. However, training such deep networks
requires a high volume of labeled data, which is not trivial to obtain in the
context of radio astronomy. Since data needs to be manually labeled by experts,
this process is not scalable to large dataset sizes, limiting the possibilities
of leveraging deep networks to address several tasks. In this work, we propose
RADiff, a generative approach based on conditional diffusion models trained
over an annotated radio dataset to generate synthetic images, containing radio
sources of different morphologies, to augment existing datasets and reduce the
problems caused by class imbalances. We also show that it is possible to
generate fully-synthetic image-annotation pairs to automatically augment any
annotated dataset. We evaluate the effectiveness of this approach by training a
semantic segmentation model on a real dataset augmented in two ways: 1) using
synthetic images obtained from real masks, and 2) generating images from
synthetic semantic masks. We show an improvement in performance when applying
augmentation, gaining up to 18% in performance when using real masks and 4%
when augmenting with synthetic masks. Finally, we employ this model to generate
large-scale radio maps with the objective of simulating Data Challenges.
| [
{
"version": "v1",
"created": "Wed, 5 Jul 2023 16:04:44 GMT"
}
] | 2023-07-06T00:00:00 | [
[
"Sortino",
"Renato",
""
],
[
"Cecconello",
"Thomas",
""
],
[
"DeMarco",
"Andrea",
""
],
[
"Fiameni",
"Giuseppe",
""
],
[
"Pilzer",
"Andrea",
""
],
[
"Hopkins",
"Andrew M.",
""
],
[
"Magro",
"Daniel",
""
],
[
"Riggi",
"Simone",
""
],
[
"Sciacca",
"Eva",
""
],
[
"Ingallinera",
"Adriano",
""
],
[
"Bordiu",
"Cristobal",
""
],
[
"Bufano",
"Filomena",
""
],
[
"Spampinato",
"Concetto",
""
]
] | TITLE: RADiff: Controllable Diffusion Models for Radio Astronomical Maps
Generation
ABSTRACT: Along with the nearing completion of the Square Kilometre Array (SKA), comes
an increasing demand for accurate and reliable automated solutions to extract
valuable information from the vast amount of data it will allow acquiring.
Automated source finding is a particularly important task in this context, as
it enables the detection and classification of astronomical objects.
Deep-learning-based object detection and semantic segmentation models have
proven to be suitable for this purpose. However, training such deep networks
requires a high volume of labeled data, which is not trivial to obtain in the
context of radio astronomy. Since data needs to be manually labeled by experts,
this process is not scalable to large dataset sizes, limiting the possibilities
of leveraging deep networks to address several tasks. In this work, we propose
RADiff, a generative approach based on conditional diffusion models trained
over an annotated radio dataset to generate synthetic images, containing radio
sources of different morphologies, to augment existing datasets and reduce the
problems caused by class imbalances. We also show that it is possible to
generate fully-synthetic image-annotation pairs to automatically augment any
annotated dataset. We evaluate the effectiveness of this approach by training a
semantic segmentation model on a real dataset augmented in two ways: 1) using
synthetic images obtained from real masks, and 2) generating images from
synthetic semantic masks. We show an improvement in performance when applying
augmentation, gaining up to 18% in performance when using real masks and 4%
when augmenting with synthetic masks. Finally, we employ this model to generate
large-scale radio maps with the objective of simulating Data Challenges.
| no_new_dataset | 0.712207 |
2307.02394 | Shengding Hu | Shengding Hu, Yifan Luo, Huadong Wang, Xingyi Cheng, Zhiyuan Liu,
Maosong Sun | Won't Get Fooled Again: Answering Questions with False Premises | Accepted to ACL 2023 main conference | null | null | null | cs.CL | http://creativecommons.org/licenses/by/4.0/ | Pre-trained language models (PLMs) have shown unprecedented potential in
various fields, especially as the backbones for question-answering (QA)
systems. However, they tend to be easily deceived by tricky questions such as
"How many eyes does the sun have?". Such frailties of PLMs often allude to the
lack of knowledge within them. In this paper, we find that the PLMs already
possess the knowledge required to rebut such questions, and the key is how to
activate the knowledge. To systematize this observation, we investigate the
PLMs' responses to one kind of tricky questions, i.e., the false premises
questions (FPQs). We annotate a FalseQA dataset containing 2365 human-written
FPQs, with the corresponding explanations for the false premises and the
revised true premise questions. Using FalseQA, we discover that PLMs are
capable of discriminating FPQs by fine-tuning on moderate numbers (e.g., 256)
of examples. PLMs also generate reasonable explanations for the false premise,
which serve as rebuttals. Further replaying a few general questions during
training allows PLMs to excel on FPQs and general questions simultaneously. Our
work suggests that once the rebuttal ability is stimulated, knowledge inside
the PLMs can be effectively utilized to handle FPQs, which incentivizes the
research on PLM-based QA systems.
| [
{
"version": "v1",
"created": "Wed, 5 Jul 2023 16:09:21 GMT"
}
] | 2023-07-06T00:00:00 | [
[
"Hu",
"Shengding",
""
],
[
"Luo",
"Yifan",
""
],
[
"Wang",
"Huadong",
""
],
[
"Cheng",
"Xingyi",
""
],
[
"Liu",
"Zhiyuan",
""
],
[
"Sun",
"Maosong",
""
]
] | TITLE: Won't Get Fooled Again: Answering Questions with False Premises
ABSTRACT: Pre-trained language models (PLMs) have shown unprecedented potential in
various fields, especially as the backbones for question-answering (QA)
systems. However, they tend to be easily deceived by tricky questions such as
"How many eyes does the sun have?". Such frailties of PLMs often allude to the
lack of knowledge within them. In this paper, we find that the PLMs already
possess the knowledge required to rebut such questions, and the key is how to
activate the knowledge. To systematize this observation, we investigate the
PLMs' responses to one kind of tricky questions, i.e., the false premises
questions (FPQs). We annotate a FalseQA dataset containing 2365 human-written
FPQs, with the corresponding explanations for the false premises and the
revised true premise questions. Using FalseQA, we discover that PLMs are
capable of discriminating FPQs by fine-tuning on moderate numbers (e.g., 256)
of examples. PLMs also generate reasonable explanations for the false premise,
which serve as rebuttals. Further replaying a few general questions during
training allows PLMs to excel on FPQs and general questions simultaneously. Our
work suggests that once the rebuttal ability is stimulated, knowledge inside
the PLMs can be effectively utilized to handle FPQs, which incentivizes the
research on PLM-based QA systems.
| new_dataset | 0.712589 |
2307.02416 | Satyajit Ghosh | Satyajit Ghosh and Mousumi Dutta | Indriya: Building a Secure and Transparent Organ Donation System with
Hyperledger Fabric | 13 pages, 4 figures, 4 tables | null | 10.36227/techrxiv.22225999.v1 | null | cs.DC | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Recent technological advancements have led to the development of new methods
for managing organ donation systems, which aim to overcome the limitations of
traditional centralized systems. To achieve increased transparency, security,
and efficiency in the organ donation process, blockchain technology is being
proposed as a replacement for these centralized systems. However, most previous
works on organ donation systems have focused on using Ethereum-based blockchain
solutions, which offer limited control, a fixed set of consensus protocols, and
no support for concurrent executions. In contrast, our work has utilized the
Hyperledger Fabric framework to develop a network model of the organ donation
system. We have designed and deployed a prototype system with smart contracts
using Amazon Managed Blockchain Service. Additionally, we have built a client
application that uses the Fabric SDK to interact with the network and perform
various actions. To evaluate the performance of our system, we conducted
extensive testing using the Hyperledger Caliper benchmarking tool. In our test
bench, the system achieved a peak actual send rate of 389.1 transactions per
second (TPS) for creating new records and 508.4 TPS for reading records. At a
send rate of 800 TPS, the system took an average of 12.16 seconds to serve a
request for creating a record and an average of 3.71 seconds to serve a request
for reading a record. Future work is required to extend the functionalities of
the system and identify potential endorsers and managers for this type of
controlled blockchain network.
| [
{
"version": "v1",
"created": "Tue, 7 Mar 2023 15:03:17 GMT"
}
] | 2023-07-06T00:00:00 | [
[
"Ghosh",
"Satyajit",
""
],
[
"Dutta",
"Mousumi",
""
]
] | TITLE: Indriya: Building a Secure and Transparent Organ Donation System with
Hyperledger Fabric
ABSTRACT: Recent technological advancements have led to the development of new methods
for managing organ donation systems, which aim to overcome the limitations of
traditional centralized systems. To achieve increased transparency, security,
and efficiency in the organ donation process, blockchain technology is being
proposed as a replacement for these centralized systems. However, most previous
works on organ donation systems have focused on using Ethereum-based blockchain
solutions, which offer limited control, a fixed set of consensus protocols, and
no support for concurrent executions. In contrast, our work has utilized the
Hyperledger Fabric framework to develop a network model of the organ donation
system. We have designed and deployed a prototype system with smart contracts
using Amazon Managed Blockchain Service. Additionally, we have built a client
application that uses the Fabric SDK to interact with the network and perform
various actions. To evaluate the performance of our system, we conducted
extensive testing using the Hyperledger Caliper benchmarking tool. In our test
bench, the system achieved a peak actual send rate of 389.1 transactions per
second (TPS) for creating new records and 508.4 TPS for reading records. At a
send rate of 800 TPS, the system took an average of 12.16 seconds to serve a
request for creating a record and an average of 3.71 seconds to serve a request
for reading a record. Future work is required to extend the functionalities of
the system and identify potential endorsers and managers for this type of
controlled blockchain network.
| no_new_dataset | 0.707436 |
2307.02435 | Prateek Yadav | Prateek Yadav, Qing Sun, Hantian Ding, Xiaopeng Li, Dejiao Zhang, Ming
Tan, Xiaofei Ma, Parminder Bhatia, Ramesh Nallapati, Murali Krishna
Ramanathan, Mohit Bansal, Bing Xiang | Exploring Continual Learning for Code Generation Models | ACL 2023 | null | null | null | cs.LG cs.CL cs.SE | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Large-scale code generation models such as Codex and CodeT5 have achieved
impressive performance. However, libraries are upgraded or deprecated very
frequently and re-training large-scale language models is computationally
expensive. Therefore, Continual Learning (CL) is an important aspect that
remains underexplored in the code domain. In this paper, we introduce a
benchmark called CodeTask-CL that covers a wide range of tasks, including code
generation, translation, summarization, and refinement, with different input
and output programming languages. Next, on our CodeTask-CL benchmark, we
compare popular CL techniques from NLP and Vision domains. We find that
effective methods like Prompt Pooling (PP) suffer from catastrophic forgetting
due to the unstable training of the prompt selection mechanism caused by stark
distribution shifts in coding tasks. We address this issue with our proposed
method, Prompt Pooling with Teacher Forcing (PP-TF), that stabilizes training
by enforcing constraints on the prompt selection mechanism and leads to a
21.54% improvement over Prompt Pooling. Along with the benchmark, we establish
a training pipeline that can be used for CL on code models, which we believe
can motivate further development of CL methods for code models. Our code is
available at https://github.com/amazon-science/codetaskcl-pptf
| [
{
"version": "v1",
"created": "Wed, 5 Jul 2023 16:58:39 GMT"
}
] | 2023-07-06T00:00:00 | [
[
"Yadav",
"Prateek",
""
],
[
"Sun",
"Qing",
""
],
[
"Ding",
"Hantian",
""
],
[
"Li",
"Xiaopeng",
""
],
[
"Zhang",
"Dejiao",
""
],
[
"Tan",
"Ming",
""
],
[
"Ma",
"Xiaofei",
""
],
[
"Bhatia",
"Parminder",
""
],
[
"Nallapati",
"Ramesh",
""
],
[
"Ramanathan",
"Murali Krishna",
""
],
[
"Bansal",
"Mohit",
""
],
[
"Xiang",
"Bing",
""
]
] | TITLE: Exploring Continual Learning for Code Generation Models
ABSTRACT: Large-scale code generation models such as Codex and CodeT5 have achieved
impressive performance. However, libraries are upgraded or deprecated very
frequently and re-training large-scale language models is computationally
expensive. Therefore, Continual Learning (CL) is an important aspect that
remains underexplored in the code domain. In this paper, we introduce a
benchmark called CodeTask-CL that covers a wide range of tasks, including code
generation, translation, summarization, and refinement, with different input
and output programming languages. Next, on our CodeTask-CL benchmark, we
compare popular CL techniques from NLP and Vision domains. We find that
effective methods like Prompt Pooling (PP) suffer from catastrophic forgetting
due to the unstable training of the prompt selection mechanism caused by stark
distribution shifts in coding tasks. We address this issue with our proposed
method, Prompt Pooling with Teacher Forcing (PP-TF), that stabilizes training
by enforcing constraints on the prompt selection mechanism and leads to a
21.54% improvement over Prompt Pooling. Along with the benchmark, we establish
a training pipeline that can be used for CL on code models, which we believe
can motivate further development of CL methods for code models. Our code is
available at https://github.com/amazon-science/codetaskcl-pptf
| no_new_dataset | 0.66405 |
2307.02464 | Ao Cheng | Ao Cheng and Guoqiang Zhao and Lirong Wang and Ruobing Zhang | AxonCallosumEM Dataset: Axon Semantic Segmentation of Whole Corpus
Callosum cross section from EM Images | 9 pages, 6 figures | null | null | null | eess.IV cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The electron microscope (EM) remains the predominant technique for
elucidating intricate details of the animal nervous system at the nanometer
scale. However, accurately reconstructing the complex morphology of axons and
myelin sheaths poses a significant challenge. Furthermore, the absence of
publicly available, large-scale EM datasets encompassing complete cross
sections of the corpus callosum, with dense ground truth segmentation for axons
and myelin sheaths, hinders the advancement and evaluation of holistic corpus
callosum reconstructions. To surmount these obstacles, we introduce the
AxonCallosumEM dataset, comprising a 1.83 times 5.76mm EM image captured from
the corpus callosum of the Rett Syndrome (RTT) mouse model, which entail
extensive axon bundles. We meticulously proofread over 600,000 patches at a
resolution of 1024 times 1024, thus providing a comprehensive ground truth for
myelinated axons and myelin sheaths. Additionally, we extensively annotated
three distinct regions within the dataset for the purposes of training,
testing, and validation. Utilizing this dataset, we develop a fine-tuning
methodology that adapts Segment Anything Model (SAM) to EM images segmentation
tasks, called EM-SAM, enabling outperforms other state-of-the-art methods.
Furthermore, we present the evaluation results of EM-SAM as a baseline.
| [
{
"version": "v1",
"created": "Wed, 5 Jul 2023 17:38:01 GMT"
}
] | 2023-07-06T00:00:00 | [
[
"Cheng",
"Ao",
""
],
[
"Zhao",
"Guoqiang",
""
],
[
"Wang",
"Lirong",
""
],
[
"Zhang",
"Ruobing",
""
]
] | TITLE: AxonCallosumEM Dataset: Axon Semantic Segmentation of Whole Corpus
Callosum cross section from EM Images
ABSTRACT: The electron microscope (EM) remains the predominant technique for
elucidating intricate details of the animal nervous system at the nanometer
scale. However, accurately reconstructing the complex morphology of axons and
myelin sheaths poses a significant challenge. Furthermore, the absence of
publicly available, large-scale EM datasets encompassing complete cross
sections of the corpus callosum, with dense ground truth segmentation for axons
and myelin sheaths, hinders the advancement and evaluation of holistic corpus
callosum reconstructions. To surmount these obstacles, we introduce the
AxonCallosumEM dataset, comprising a 1.83 times 5.76mm EM image captured from
the corpus callosum of the Rett Syndrome (RTT) mouse model, which entail
extensive axon bundles. We meticulously proofread over 600,000 patches at a
resolution of 1024 times 1024, thus providing a comprehensive ground truth for
myelinated axons and myelin sheaths. Additionally, we extensively annotated
three distinct regions within the dataset for the purposes of training,
testing, and validation. Utilizing this dataset, we develop a fine-tuning
methodology that adapts Segment Anything Model (SAM) to EM images segmentation
tasks, called EM-SAM, enabling outperforms other state-of-the-art methods.
Furthermore, we present the evaluation results of EM-SAM as a baseline.
| new_dataset | 0.716175 |
2307.02465 | Marc Ru{\ss}wurm | Marc Ru{\ss}wurm, Sushen Jilla Venkatesa, Devis Tuia | Large-scale Detection of Marine Debris in Coastal Areas with Sentinel-2 | in review | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | Detecting and quantifying marine pollution and macro-plastics is an
increasingly pressing ecological issue that directly impacts ecology and human
health. Efforts to quantify marine pollution are often conducted with sparse
and expensive beach surveys, which are difficult to conduct on a large scale.
Here, remote sensing can provide reliable estimates of plastic pollution by
regularly monitoring and detecting marine debris in coastal areas.
Medium-resolution satellite data of coastal areas is readily available and can
be leveraged to detect aggregations of marine debris containing plastic litter.
In this work, we present a detector for marine debris built on a deep
segmentation model that outputs a probability for marine debris at the pixel
level. We train this detector with a combination of annotated datasets of
marine debris and evaluate it on specifically selected test sites where it is
highly probable that plastic pollution is present in the detected marine
debris. We demonstrate quantitatively and qualitatively that a deep learning
model trained on this dataset issued from multiple sources outperforms existing
detection models trained on previous datasets by a large margin. Our
experiments show, consistent with the principles of data-centric AI, that this
performance is due to our particular dataset design with extensive sampling of
negative examples and label refinements rather than depending on the particular
deep learning model. We hope to accelerate advances in the large-scale
automated detection of marine debris, which is a step towards quantifying and
monitoring marine litter with remote sensing at global scales, and release the
model weights and training source code under
https://github.com/marccoru/marinedebrisdetector
| [
{
"version": "v1",
"created": "Wed, 5 Jul 2023 17:38:48 GMT"
}
] | 2023-07-06T00:00:00 | [
[
"Rußwurm",
"Marc",
""
],
[
"Venkatesa",
"Sushen Jilla",
""
],
[
"Tuia",
"Devis",
""
]
] | TITLE: Large-scale Detection of Marine Debris in Coastal Areas with Sentinel-2
ABSTRACT: Detecting and quantifying marine pollution and macro-plastics is an
increasingly pressing ecological issue that directly impacts ecology and human
health. Efforts to quantify marine pollution are often conducted with sparse
and expensive beach surveys, which are difficult to conduct on a large scale.
Here, remote sensing can provide reliable estimates of plastic pollution by
regularly monitoring and detecting marine debris in coastal areas.
Medium-resolution satellite data of coastal areas is readily available and can
be leveraged to detect aggregations of marine debris containing plastic litter.
In this work, we present a detector for marine debris built on a deep
segmentation model that outputs a probability for marine debris at the pixel
level. We train this detector with a combination of annotated datasets of
marine debris and evaluate it on specifically selected test sites where it is
highly probable that plastic pollution is present in the detected marine
debris. We demonstrate quantitatively and qualitatively that a deep learning
model trained on this dataset issued from multiple sources outperforms existing
detection models trained on previous datasets by a large margin. Our
experiments show, consistent with the principles of data-centric AI, that this
performance is due to our particular dataset design with extensive sampling of
negative examples and label refinements rather than depending on the particular
deep learning model. We hope to accelerate advances in the large-scale
automated detection of marine debris, which is a step towards quantifying and
monitoring marine litter with remote sensing at global scales, and release the
model weights and training source code under
https://github.com/marccoru/marinedebrisdetector
| new_dataset | 0.699421 |
2307.02480 | Hari Gupta | Hari Prabhat Gupta and Rahul Mishra | A Dataset of Inertial Measurement Units for Handwritten English
Alphabets | 10 pages, 12 figures | null | 10.21227/av6q-jj17 | null | cs.CV | http://creativecommons.org/licenses/by-nc-nd/4.0/ | This paper presents an end-to-end methodology for collecting datasets to
recognize handwritten English alphabets by utilizing Inertial Measurement Units
(IMUs) and leveraging the diversity present in the Indian writing style. The
IMUs are utilized to capture the dynamic movement patterns associated with
handwriting, enabling more accurate recognition of alphabets. The Indian
context introduces various challenges due to the heterogeneity in writing
styles across different regions and languages. By leveraging this diversity,
the collected dataset and the collection system aim to achieve higher
recognition accuracy. Some preliminary experimental results demonstrate the
effectiveness of the dataset in accurately recognizing handwritten English
alphabet in the Indian context. This research can be extended and contributes
to the field of pattern recognition and offers valuable insights for developing
improved systems for handwriting recognition, particularly in diverse
linguistic and cultural contexts.
| [
{
"version": "v1",
"created": "Wed, 5 Jul 2023 17:54:36 GMT"
}
] | 2023-07-06T00:00:00 | [
[
"Gupta",
"Hari Prabhat",
""
],
[
"Mishra",
"Rahul",
""
]
] | TITLE: A Dataset of Inertial Measurement Units for Handwritten English
Alphabets
ABSTRACT: This paper presents an end-to-end methodology for collecting datasets to
recognize handwritten English alphabets by utilizing Inertial Measurement Units
(IMUs) and leveraging the diversity present in the Indian writing style. The
IMUs are utilized to capture the dynamic movement patterns associated with
handwriting, enabling more accurate recognition of alphabets. The Indian
context introduces various challenges due to the heterogeneity in writing
styles across different regions and languages. By leveraging this diversity,
the collected dataset and the collection system aim to achieve higher
recognition accuracy. Some preliminary experimental results demonstrate the
effectiveness of the dataset in accurately recognizing handwritten English
alphabet in the Indian context. This research can be extended and contributes
to the field of pattern recognition and offers valuable insights for developing
improved systems for handwriting recognition, particularly in diverse
linguistic and cultural contexts.
| no_new_dataset | 0.552057 |
2111.11843 | Liheng Bian | Lintao Peng, Chunli Zhu, Liheng Bian | U-shape Transformer for Underwater Image Enhancement | under review | null | 10.1109/TIP.2023.3276332 | null | cs.CV eess.IV | http://creativecommons.org/licenses/by/4.0/ | The light absorption and scattering of underwater impurities lead to poor
underwater imaging quality. The existing data-driven based underwater image
enhancement (UIE) techniques suffer from the lack of a large-scale dataset
containing various underwater scenes and high-fidelity reference images.
Besides, the inconsistent attenuation in different color channels and space
areas is not fully considered for boosted enhancement. In this work, we
constructed a large-scale underwater image (LSUI) dataset including 5004 image
pairs, and reported an U-shape Transformer network where the transformer model
is for the first time introduced to the UIE task. The U-shape Transformer is
integrated with a channel-wise multi-scale feature fusion transformer (CMSFFT)
module and a spatial-wise global feature modeling transformer (SGFMT) module,
which reinforce the network's attention to the color channels and space areas
with more serious attenuation. Meanwhile, in order to further improve the
contrast and saturation, a novel loss function combining RGB, LAB and LCH color
spaces is designed following the human vision principle. The extensive
experiments on available datasets validate the state-of-the-art performance of
the reported technique with more than 2dB superiority.
| [
{
"version": "v1",
"created": "Tue, 23 Nov 2021 13:15:56 GMT"
},
{
"version": "v2",
"created": "Wed, 24 Nov 2021 04:49:34 GMT"
},
{
"version": "v3",
"created": "Fri, 3 Dec 2021 05:44:54 GMT"
},
{
"version": "v4",
"created": "Wed, 23 Mar 2022 11:07:28 GMT"
},
{
"version": "v5",
"created": "Wed, 4 May 2022 14:19:46 GMT"
},
{
"version": "v6",
"created": "Sun, 12 Jun 2022 11:45:40 GMT"
}
] | 2023-07-05T00:00:00 | [
[
"Peng",
"Lintao",
""
],
[
"Zhu",
"Chunli",
""
],
[
"Bian",
"Liheng",
""
]
] | TITLE: U-shape Transformer for Underwater Image Enhancement
ABSTRACT: The light absorption and scattering of underwater impurities lead to poor
underwater imaging quality. The existing data-driven based underwater image
enhancement (UIE) techniques suffer from the lack of a large-scale dataset
containing various underwater scenes and high-fidelity reference images.
Besides, the inconsistent attenuation in different color channels and space
areas is not fully considered for boosted enhancement. In this work, we
constructed a large-scale underwater image (LSUI) dataset including 5004 image
pairs, and reported an U-shape Transformer network where the transformer model
is for the first time introduced to the UIE task. The U-shape Transformer is
integrated with a channel-wise multi-scale feature fusion transformer (CMSFFT)
module and a spatial-wise global feature modeling transformer (SGFMT) module,
which reinforce the network's attention to the color channels and space areas
with more serious attenuation. Meanwhile, in order to further improve the
contrast and saturation, a novel loss function combining RGB, LAB and LCH color
spaces is designed following the human vision principle. The extensive
experiments on available datasets validate the state-of-the-art performance of
the reported technique with more than 2dB superiority.
| new_dataset | 0.719799 |
2205.04126 | Yueying Kao | Yueying Kao and Bowen Pan and Miao Xu and Jiangjing Lyu and Xiangyu
Zhu and Yuanzhang Chang and Xiaobo Li and Zhen Lei | Towards 3D Face Reconstruction in Perspective Projection: Estimating
6DoF Face Pose from Monocular Image | Accepted by TIP | null | 10.1109/TIP.2023.3275535 | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In 3D face reconstruction, orthogonal projection has been widely employed to
substitute perspective projection to simplify the fitting process. This
approximation performs well when the distance between camera and face is far
enough. However, in some scenarios that the face is very close to camera or
moving along the camera axis, the methods suffer from the inaccurate
reconstruction and unstable temporal fitting due to the distortion under the
perspective projection. In this paper, we aim to address the problem of
single-image 3D face reconstruction under perspective projection. Specifically,
a deep neural network, Perspective Network (PerspNet), is proposed to
simultaneously reconstruct 3D face shape in canonical space and learn the
correspondence between 2D pixels and 3D points, by which the 6DoF (6 Degrees of
Freedom) face pose can be estimated to represent perspective projection.
Besides, we contribute a large ARKitFace dataset to enable the training and
evaluation of 3D face reconstruction solutions under the scenarios of
perspective projection, which has 902,724 2D facial images with ground-truth 3D
face mesh and annotated 6DoF pose parameters. Experimental results show that
our approach outperforms current state-of-the-art methods by a significant
margin. The code and data are available at
https://github.com/cbsropenproject/6dof_face.
| [
{
"version": "v1",
"created": "Mon, 9 May 2022 08:49:41 GMT"
},
{
"version": "v2",
"created": "Wed, 17 May 2023 11:35:41 GMT"
}
] | 2023-07-05T00:00:00 | [
[
"Kao",
"Yueying",
""
],
[
"Pan",
"Bowen",
""
],
[
"Xu",
"Miao",
""
],
[
"Lyu",
"Jiangjing",
""
],
[
"Zhu",
"Xiangyu",
""
],
[
"Chang",
"Yuanzhang",
""
],
[
"Li",
"Xiaobo",
""
],
[
"Lei",
"Zhen",
""
]
] | TITLE: Towards 3D Face Reconstruction in Perspective Projection: Estimating
6DoF Face Pose from Monocular Image
ABSTRACT: In 3D face reconstruction, orthogonal projection has been widely employed to
substitute perspective projection to simplify the fitting process. This
approximation performs well when the distance between camera and face is far
enough. However, in some scenarios that the face is very close to camera or
moving along the camera axis, the methods suffer from the inaccurate
reconstruction and unstable temporal fitting due to the distortion under the
perspective projection. In this paper, we aim to address the problem of
single-image 3D face reconstruction under perspective projection. Specifically,
a deep neural network, Perspective Network (PerspNet), is proposed to
simultaneously reconstruct 3D face shape in canonical space and learn the
correspondence between 2D pixels and 3D points, by which the 6DoF (6 Degrees of
Freedom) face pose can be estimated to represent perspective projection.
Besides, we contribute a large ARKitFace dataset to enable the training and
evaluation of 3D face reconstruction solutions under the scenarios of
perspective projection, which has 902,724 2D facial images with ground-truth 3D
face mesh and annotated 6DoF pose parameters. Experimental results show that
our approach outperforms current state-of-the-art methods by a significant
margin. The code and data are available at
https://github.com/cbsropenproject/6dof_face.
| new_dataset | 0.71682 |
2206.06427 | Zhenyu Wu | Priya Narayanan, Xin Hu, Zhenyu Wu, Matthew D Thielke, John G Rogers,
Andre V Harrison, John A D'Agostino, James D Brown, Long P Quang, James R
Uplinger, Heesung Kwon, Zhangyang Wang | A Multi-purpose Realistic Haze Benchmark with Quantifiable Haze Levels
and Ground Truth | This paper has been ACCEPTED for publication as a REGULAR paper in
the IEEE Transactions on Image Processing (TIP) | null | 10.1109/TIP.2023.3245994 | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Imagery collected from outdoor visual environments is often degraded due to
the presence of dense smoke or haze. A key challenge for research in scene
understanding in these degraded visual environments (DVE) is the lack of
representative benchmark datasets. These datasets are required to evaluate
state-of-the-art vision algorithms (e.g., detection and tracking) in degraded
settings. In this paper, we address some of these limitations by introducing
the first realistic hazy image benchmark, from both aerial and ground view,
with paired haze-free images, and in-situ haze density measurements. This
dataset was produced in a controlled environment with professional smoke
generating machines that covered the entire scene, and consists of images
captured from the perspective of both an unmanned aerial vehicle (UAV) and an
unmanned ground vehicle (UGV). We also evaluate a set of representative
state-of-the-art dehazing approaches as well as object detectors on the
dataset. The full dataset presented in this paper, including the ground truth
object classification bounding boxes and haze density measurements, is provided
for the community to evaluate their algorithms at:
https://a2i2-archangel.vision. A subset of this dataset has been used for the
``Object Detection in Haze'' Track of CVPR UG2 2022 challenge at
http://cvpr2022.ug2challenge.org/track1.html.
| [
{
"version": "v1",
"created": "Mon, 13 Jun 2022 19:14:06 GMT"
},
{
"version": "v2",
"created": "Sun, 3 Jul 2022 01:09:44 GMT"
},
{
"version": "v3",
"created": "Sat, 11 Feb 2023 20:36:36 GMT"
}
] | 2023-07-05T00:00:00 | [
[
"Narayanan",
"Priya",
""
],
[
"Hu",
"Xin",
""
],
[
"Wu",
"Zhenyu",
""
],
[
"Thielke",
"Matthew D",
""
],
[
"Rogers",
"John G",
""
],
[
"Harrison",
"Andre V",
""
],
[
"D'Agostino",
"John A",
""
],
[
"Brown",
"James D",
""
],
[
"Quang",
"Long P",
""
],
[
"Uplinger",
"James R",
""
],
[
"Kwon",
"Heesung",
""
],
[
"Wang",
"Zhangyang",
""
]
] | TITLE: A Multi-purpose Realistic Haze Benchmark with Quantifiable Haze Levels
and Ground Truth
ABSTRACT: Imagery collected from outdoor visual environments is often degraded due to
the presence of dense smoke or haze. A key challenge for research in scene
understanding in these degraded visual environments (DVE) is the lack of
representative benchmark datasets. These datasets are required to evaluate
state-of-the-art vision algorithms (e.g., detection and tracking) in degraded
settings. In this paper, we address some of these limitations by introducing
the first realistic hazy image benchmark, from both aerial and ground view,
with paired haze-free images, and in-situ haze density measurements. This
dataset was produced in a controlled environment with professional smoke
generating machines that covered the entire scene, and consists of images
captured from the perspective of both an unmanned aerial vehicle (UAV) and an
unmanned ground vehicle (UGV). We also evaluate a set of representative
state-of-the-art dehazing approaches as well as object detectors on the
dataset. The full dataset presented in this paper, including the ground truth
object classification bounding boxes and haze density measurements, is provided
for the community to evaluate their algorithms at:
https://a2i2-archangel.vision. A subset of this dataset has been used for the
``Object Detection in Haze'' Track of CVPR UG2 2022 challenge at
http://cvpr2022.ug2challenge.org/track1.html.
| new_dataset | 0.719186 |
2303.03918 | Mostafa Mobasher | Panos Pantidis, Habiba Eldababy, Christopher Miguel Tagle, Mostafa E.
Mobasher | Error convergence and engineering-guided hyperparameter search of PINNs:
towards optimized I-FENN performance | null | null | 10.1016/j.cma.2023.116160 | null | cs.LG | http://creativecommons.org/licenses/by-nc-nd/4.0/ | In our recently proposed Integrated Finite Element Neural Network (I-FENN)
framework (Pantidis and Mobasher, 2023) we showcased how PINNs can be deployed
on a finite element-level basis to swiftly approximate a state variable of
interest, and we applied it in the context of non-local gradient-enhanced
damage mechanics. In this paper, we enhance the rigour and performance of
I-FENN by focusing on two crucial aspects of its PINN component: a) the error
convergence analysis and b) the hyperparameter-performance relationship. Guided
by the available theoretical formulations in the field, we introduce a
systematic numerical approach based on a novel set of holistic performance
metrics to answer both objectives. In the first objective, we explore in detail
the convergence of the PINN training error and the global error against the
network size and the training sample size. We demonstrate a consistent
converging behavior of the two error types for any investigated combination of
network complexity, dataset size and choice of hyperparameters, which
empirically proves the conformance of the PINN setup and implementation to the
available convergence theories. In the second objective, we establish an
a-priori knowledge of the hyperparameters which favor higher predictive
accuracy, lower computational effort, and the least chances of arriving at
trivial solutions. The analysis leads to several outcomes that contribute to
the better performance of I-FENN, and fills a long-standing gap in the PINN
literature with regards to the numerical convergence of the network errors
while accounting for commonly used optimizers (Adam and L-BFGS). The proposed
analysis method can be directly extended to other ML applications in science
and engineering. The code and data utilized in the analysis are posted publicly
to aid the reproduction and extension of this research.
| [
{
"version": "v1",
"created": "Fri, 3 Mar 2023 17:39:06 GMT"
},
{
"version": "v2",
"created": "Mon, 5 Jun 2023 07:40:27 GMT"
}
] | 2023-07-05T00:00:00 | [
[
"Pantidis",
"Panos",
""
],
[
"Eldababy",
"Habiba",
""
],
[
"Tagle",
"Christopher Miguel",
""
],
[
"Mobasher",
"Mostafa E.",
""
]
] | TITLE: Error convergence and engineering-guided hyperparameter search of PINNs:
towards optimized I-FENN performance
ABSTRACT: In our recently proposed Integrated Finite Element Neural Network (I-FENN)
framework (Pantidis and Mobasher, 2023) we showcased how PINNs can be deployed
on a finite element-level basis to swiftly approximate a state variable of
interest, and we applied it in the context of non-local gradient-enhanced
damage mechanics. In this paper, we enhance the rigour and performance of
I-FENN by focusing on two crucial aspects of its PINN component: a) the error
convergence analysis and b) the hyperparameter-performance relationship. Guided
by the available theoretical formulations in the field, we introduce a
systematic numerical approach based on a novel set of holistic performance
metrics to answer both objectives. In the first objective, we explore in detail
the convergence of the PINN training error and the global error against the
network size and the training sample size. We demonstrate a consistent
converging behavior of the two error types for any investigated combination of
network complexity, dataset size and choice of hyperparameters, which
empirically proves the conformance of the PINN setup and implementation to the
available convergence theories. In the second objective, we establish an
a-priori knowledge of the hyperparameters which favor higher predictive
accuracy, lower computational effort, and the least chances of arriving at
trivial solutions. The analysis leads to several outcomes that contribute to
the better performance of I-FENN, and fills a long-standing gap in the PINN
literature with regards to the numerical convergence of the network errors
while accounting for commonly used optimizers (Adam and L-BFGS). The proposed
analysis method can be directly extended to other ML applications in science
and engineering. The code and data utilized in the analysis are posted publicly
to aid the reproduction and extension of this research.
| no_new_dataset | 0.709466 |
2303.07292 | Olumide Ebenezer Ojo | Olumide Ebenezer Ojo, Hoang Thang Ta, Alexander Gelbukh, Hiram Calvo,
Olaronke Oluwayemisi Adebanji, Grigori Sidorov | Transformer-based approaches to Sentiment Detection | This submission has been removed from arXiv because the submitter did
not have the authority to grant the license at the time of submission | null | null | null | cs.CL | http://creativecommons.org/licenses/by/4.0/ | The use of transfer learning methods is largely responsible for the present
breakthrough in Natural Learning Processing (NLP) tasks across multiple
domains. In order to solve the problem of sentiment detection, we examined the
performance of four different types of well-known state-of-the-art transformer
models for text classification. Models such as Bidirectional Encoder
Representations from Transformers (BERT), Robustly Optimized BERT Pre-training
Approach (RoBERTa), a distilled version of BERT (DistilBERT), and a large
bidirectional neural network architecture (XLNet) were proposed. The
performance of the four models that were used to detect disaster in the text
was compared. All the models performed well enough, indicating that
transformer-based models are suitable for the detection of disaster in text.
The RoBERTa transformer model performs best on the test dataset with a score of
82.6% and is highly recommended for quality predictions. Furthermore, we
discovered that the learning algorithms' performance was influenced by the
pre-processing techniques, the nature of words in the vocabulary, unbalanced
labeling, and the model parameters.
| [
{
"version": "v1",
"created": "Mon, 13 Mar 2023 17:12:03 GMT"
}
] | 2023-07-05T00:00:00 | [
[
"Ojo",
"Olumide Ebenezer",
""
],
[
"Ta",
"Hoang Thang",
""
],
[
"Gelbukh",
"Alexander",
""
],
[
"Calvo",
"Hiram",
""
],
[
"Adebanji",
"Olaronke Oluwayemisi",
""
],
[
"Sidorov",
"Grigori",
""
]
] | TITLE: Transformer-based approaches to Sentiment Detection
ABSTRACT: The use of transfer learning methods is largely responsible for the present
breakthrough in Natural Learning Processing (NLP) tasks across multiple
domains. In order to solve the problem of sentiment detection, we examined the
performance of four different types of well-known state-of-the-art transformer
models for text classification. Models such as Bidirectional Encoder
Representations from Transformers (BERT), Robustly Optimized BERT Pre-training
Approach (RoBERTa), a distilled version of BERT (DistilBERT), and a large
bidirectional neural network architecture (XLNet) were proposed. The
performance of the four models that were used to detect disaster in the text
was compared. All the models performed well enough, indicating that
transformer-based models are suitable for the detection of disaster in text.
The RoBERTa transformer model performs best on the test dataset with a score of
82.6% and is highly recommended for quality predictions. Furthermore, we
discovered that the learning algorithms' performance was influenced by the
pre-processing techniques, the nature of words in the vocabulary, unbalanced
labeling, and the model parameters.
| no_new_dataset | 0.711387 |
2306.02433 | Ye Xue | Ye Xue, Vincent Lau | Riemannian Low-Rank Model Compression for Federated Learning with
Over-the-Air Aggregation | This paper has been accepted for publication as a REGULAR paper in
the IEEE Transactions on Signal Processing | null | 10.1109/TSP.2023.3284381 | null | eess.SP cs.LG | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Low-rank model compression is a widely used technique for reducing the
computational load when training machine learning models. However, existing
methods often rely on relaxing the low-rank constraint of the model weights
using a regularized nuclear norm penalty, which requires an appropriate
hyperparameter that can be difficult to determine in practice. Furthermore,
existing compression techniques are not directly applicable to efficient
over-the-air (OTA) aggregation in federated learning (FL) systems for
distributed Internet-of-Things (IoT) scenarios. In this paper, we propose a
novel manifold optimization formulation for low-rank model compression in FL
that does not relax the low-rank constraint. Our optimization is conducted
directly over the low-rank manifold, guaranteeing that the model is exactly
low-rank. We also introduce a consensus penalty in the optimization formulation
to support OTA aggregation. Based on our optimization formulation, we propose
an alternating Riemannian optimization algorithm with a precoder that enables
efficient OTA aggregation of low-rank local models without sacrificing training
performance. Additionally, we provide convergence analysis in terms of key
system parameters and conduct extensive experiments with real-world datasets to
demonstrate the effectiveness of our proposed Riemannian low-rank model
compression scheme compared to various state-of-the-art baselines.
| [
{
"version": "v1",
"created": "Sun, 4 Jun 2023 18:32:50 GMT"
}
] | 2023-07-05T00:00:00 | [
[
"Xue",
"Ye",
""
],
[
"Lau",
"Vincent",
""
]
] | TITLE: Riemannian Low-Rank Model Compression for Federated Learning with
Over-the-Air Aggregation
ABSTRACT: Low-rank model compression is a widely used technique for reducing the
computational load when training machine learning models. However, existing
methods often rely on relaxing the low-rank constraint of the model weights
using a regularized nuclear norm penalty, which requires an appropriate
hyperparameter that can be difficult to determine in practice. Furthermore,
existing compression techniques are not directly applicable to efficient
over-the-air (OTA) aggregation in federated learning (FL) systems for
distributed Internet-of-Things (IoT) scenarios. In this paper, we propose a
novel manifold optimization formulation for low-rank model compression in FL
that does not relax the low-rank constraint. Our optimization is conducted
directly over the low-rank manifold, guaranteeing that the model is exactly
low-rank. We also introduce a consensus penalty in the optimization formulation
to support OTA aggregation. Based on our optimization formulation, we propose
an alternating Riemannian optimization algorithm with a precoder that enables
efficient OTA aggregation of low-rank local models without sacrificing training
performance. Additionally, we provide convergence analysis in terms of key
system parameters and conduct extensive experiments with real-world datasets to
demonstrate the effectiveness of our proposed Riemannian low-rank model
compression scheme compared to various state-of-the-art baselines.
| no_new_dataset | 0.710597 |
2306.03501 | Thomas Wick | Leon Maximilian Kolditz, Katrin Mang, Thomas Wick | A modified combined active-set Newton method for solving phase-field
fracture into the monolithic limit | 49 pages, 45 figures, 9 tables | null | 10.1016/j.cma.2023.116170 | null | math.NA cs.NA | http://creativecommons.org/licenses/by-nc-nd/4.0/ | In this work, we examine a numerical phase-field fracture framework in which
the crack irreversibility constraint is treated with a primal-dual active set
method and a linearization is used in the degradation function to enhance the
numerical stability. The first goal is to carefully derive from a
complementarity system our primal-dual active set formulation, which has been
used in the literature in numerous studies, but for phase-field fracture
without its detailed mathematical derivation yet. Based on the latter, we
formulate a modified combined active-set Newton approach that significantly
reduces the computational cost in comparison to comparable prior algorithms for
quasi-monolithic settings. For many practical problems, Newton converges fast,
but active set needs many iterations, for which three different efficiency
improvements are suggested in this paper. Afterwards, we design an iteration on
the linearization in order to iterate the problem to the monolithic limit. Our
new algorithms are implemented in the programming framework pfm-cracks [T.
Heister, T. Wick; pfm-cracks: A parallel-adaptive framework for phase-field
fracture propagation, Software Impacts, Vol. 6 (2020), 100045]. In the
numerical examples, we conduct performance studies and investigate efficiency
enhancements. The main emphasis is on the cost complexity by keeping the
accuracy of numerical solutions and goal functionals. Our algorithmic
suggestions are substantiated with the help of several benchmarks in two and
three spatial dimensions. Therein, predictor-corrector adaptivity and parallel
performance studies are explored as well.
| [
{
"version": "v1",
"created": "Tue, 6 Jun 2023 08:40:26 GMT"
},
{
"version": "v2",
"created": "Fri, 16 Jun 2023 16:00:23 GMT"
}
] | 2023-07-05T00:00:00 | [
[
"Kolditz",
"Leon Maximilian",
""
],
[
"Mang",
"Katrin",
""
],
[
"Wick",
"Thomas",
""
]
] | TITLE: A modified combined active-set Newton method for solving phase-field
fracture into the monolithic limit
ABSTRACT: In this work, we examine a numerical phase-field fracture framework in which
the crack irreversibility constraint is treated with a primal-dual active set
method and a linearization is used in the degradation function to enhance the
numerical stability. The first goal is to carefully derive from a
complementarity system our primal-dual active set formulation, which has been
used in the literature in numerous studies, but for phase-field fracture
without its detailed mathematical derivation yet. Based on the latter, we
formulate a modified combined active-set Newton approach that significantly
reduces the computational cost in comparison to comparable prior algorithms for
quasi-monolithic settings. For many practical problems, Newton converges fast,
but active set needs many iterations, for which three different efficiency
improvements are suggested in this paper. Afterwards, we design an iteration on
the linearization in order to iterate the problem to the monolithic limit. Our
new algorithms are implemented in the programming framework pfm-cracks [T.
Heister, T. Wick; pfm-cracks: A parallel-adaptive framework for phase-field
fracture propagation, Software Impacts, Vol. 6 (2020), 100045]. In the
numerical examples, we conduct performance studies and investigate efficiency
enhancements. The main emphasis is on the cost complexity by keeping the
accuracy of numerical solutions and goal functionals. Our algorithmic
suggestions are substantiated with the help of several benchmarks in two and
three spatial dimensions. Therein, predictor-corrector adaptivity and parallel
performance studies are explored as well.
| no_new_dataset | 0.709806 |
2306.11942 | Yogesh Bansal | Yogesh Bansal, David Lillis, Mohand Tahar Kechadi | A Deep Learning Model for Heterogeneous Dataset Analysis -- Application
to Winter Wheat Crop Yield Prediction | This version has been removed by arXiv administrators because the
submitter did not have the authority to grant the license at the time of
submission | null | null | null | cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Western countries rely heavily on wheat, and yield prediction is crucial.
Time-series deep learning models, such as Long Short Term Memory (LSTM), have
already been explored and applied to yield prediction. Existing literature
reported that they perform better than traditional Machine Learning (ML)
models. However, the existing LSTM cannot handle heterogeneous datasets (a
combination of data which varies and remains static with time). In this paper,
we propose an efficient deep learning model that can deal with heterogeneous
datasets. We developed the system architecture and applied it to the real-world
dataset in the digital agriculture area. We showed that it outperforms the
existing ML models.
| [
{
"version": "v1",
"created": "Tue, 20 Jun 2023 23:39:06 GMT"
}
] | 2023-07-05T00:00:00 | [
[
"Bansal",
"Yogesh",
""
],
[
"Lillis",
"David",
""
],
[
"Kechadi",
"Mohand Tahar",
""
]
] | TITLE: A Deep Learning Model for Heterogeneous Dataset Analysis -- Application
to Winter Wheat Crop Yield Prediction
ABSTRACT: Western countries rely heavily on wheat, and yield prediction is crucial.
Time-series deep learning models, such as Long Short Term Memory (LSTM), have
already been explored and applied to yield prediction. Existing literature
reported that they perform better than traditional Machine Learning (ML)
models. However, the existing LSTM cannot handle heterogeneous datasets (a
combination of data which varies and remains static with time). In this paper,
we propose an efficient deep learning model that can deal with heterogeneous
datasets. We developed the system architecture and applied it to the real-world
dataset in the digital agriculture area. We showed that it outperforms the
existing ML models.
| no_new_dataset | 0.710409 |
1911.07706 | Hongtao Lv | Hongtao Lv, Chaoli Zhang, Zhenzhe Zheng, Tie Luo, Fan Wu, Guihai Chen | Mechanism Design with Predicted Task Revenue for Bike Sharing Systems | Accepted by AAAI 2020; This is the full version that contains all the
proofs | null | null | null | cs.GT cs.MA | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Bike sharing systems have been widely deployed around the world in recent
years. A core problem in such systems is to reposition the bikes so that the
distribution of bike supply is reshaped to better match the dynamic bike
demand. When the bike-sharing company or platform is able to predict the
revenue of each reposition task based on historic data, an additional
constraint is to cap the payment for each task below its predicted revenue. In
this paper, we propose an incentive mechanism called {\em TruPreTar} to
incentivize users to park bicycles at locations desired by the platform toward
rebalancing supply and demand. TruPreTar possesses four important economic and
computational properties such as truthfulness and budget feasibility.
Furthermore, we prove that even when the payment budget is tight, the total
revenue still exceeds or equals the budget. Otherwise, TruPreTar achieves
2-approximation as compared to the optimal (revenue-maximizing) solution, which
is close to the lower bound of at least $\sqrt{2}$ that we also prove. Using an
industrial dataset obtained from a large bike-sharing company, our experiments
show that TruPreTar is effective in rebalancing bike supply and demand and, as
a result, generates high revenue that outperforms several benchmark mechanisms.
| [
{
"version": "v1",
"created": "Mon, 18 Nov 2019 15:34:11 GMT"
},
{
"version": "v2",
"created": "Thu, 21 Nov 2019 17:49:38 GMT"
},
{
"version": "v3",
"created": "Sun, 24 Nov 2019 08:38:07 GMT"
},
{
"version": "v4",
"created": "Mon, 3 Jul 2023 09:24:59 GMT"
}
] | 2023-07-04T00:00:00 | [
[
"Lv",
"Hongtao",
""
],
[
"Zhang",
"Chaoli",
""
],
[
"Zheng",
"Zhenzhe",
""
],
[
"Luo",
"Tie",
""
],
[
"Wu",
"Fan",
""
],
[
"Chen",
"Guihai",
""
]
] | TITLE: Mechanism Design with Predicted Task Revenue for Bike Sharing Systems
ABSTRACT: Bike sharing systems have been widely deployed around the world in recent
years. A core problem in such systems is to reposition the bikes so that the
distribution of bike supply is reshaped to better match the dynamic bike
demand. When the bike-sharing company or platform is able to predict the
revenue of each reposition task based on historic data, an additional
constraint is to cap the payment for each task below its predicted revenue. In
this paper, we propose an incentive mechanism called {\em TruPreTar} to
incentivize users to park bicycles at locations desired by the platform toward
rebalancing supply and demand. TruPreTar possesses four important economic and
computational properties such as truthfulness and budget feasibility.
Furthermore, we prove that even when the payment budget is tight, the total
revenue still exceeds or equals the budget. Otherwise, TruPreTar achieves
2-approximation as compared to the optimal (revenue-maximizing) solution, which
is close to the lower bound of at least $\sqrt{2}$ that we also prove. Using an
industrial dataset obtained from a large bike-sharing company, our experiments
show that TruPreTar is effective in rebalancing bike supply and demand and, as
a result, generates high revenue that outperforms several benchmark mechanisms.
| no_new_dataset | 0.704554 |
2002.08412 | Raed Al Kontar | Seokhyun Chung, Raed Al Kontar, Zhenke Wu | Weakly-supervised Multi-output Regression via Correlated Gaussian
Processes | null | Informs Journal on Data Science, 2022 | 10.1287/ijds.2022.0018 | null | stat.ML cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Multi-output regression seeks to borrow strength and leverage commonalities
across different but related outputs in order to enhance learning and
prediction accuracy. A fundamental assumption is that the output/group
membership labels for all observations are known. This assumption is often
violated in real applications. For instance, in healthcare datasets, sensitive
attributes such as ethnicity are often missing or unreported. To this end, we
introduce a weakly-supervised multi-output model based on dependent Gaussian
processes. Our approach is able to leverage data without complete group labels
or possibly only prior belief on group memberships to enhance accuracy across
all outputs. Through intensive simulations and case studies on an Insulin,
Testosterone and Bodyfat dataset, we show that our model excels in multi-output
settings with missing labels, while being competitive in traditional fully
labeled settings. We end by highlighting the possible use of our approach in
fair inference and sequential decision-making.
| [
{
"version": "v1",
"created": "Wed, 19 Feb 2020 19:54:25 GMT"
},
{
"version": "v2",
"created": "Mon, 23 May 2022 20:56:29 GMT"
}
] | 2023-07-04T00:00:00 | [
[
"Chung",
"Seokhyun",
""
],
[
"Kontar",
"Raed Al",
""
],
[
"Wu",
"Zhenke",
""
]
] | TITLE: Weakly-supervised Multi-output Regression via Correlated Gaussian
Processes
ABSTRACT: Multi-output regression seeks to borrow strength and leverage commonalities
across different but related outputs in order to enhance learning and
prediction accuracy. A fundamental assumption is that the output/group
membership labels for all observations are known. This assumption is often
violated in real applications. For instance, in healthcare datasets, sensitive
attributes such as ethnicity are often missing or unreported. To this end, we
introduce a weakly-supervised multi-output model based on dependent Gaussian
processes. Our approach is able to leverage data without complete group labels
or possibly only prior belief on group memberships to enhance accuracy across
all outputs. Through intensive simulations and case studies on an Insulin,
Testosterone and Bodyfat dataset, we show that our model excels in multi-output
settings with missing labels, while being competitive in traditional fully
labeled settings. We end by highlighting the possible use of our approach in
fair inference and sequential decision-making.
| no_new_dataset | 0.710377 |
2004.02382 | Raed Al Kontar | Moyan Li, Raed Kontar | On Negative Transfer and Structure of Latent Functions in Multi-output
Gaussian Processes | null | SIAM Journal on Uncertainty Quantification, 2022 | 10.1137/21M1436816 | null | stat.ML cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The multi-output Gaussian process ($\mathcal{MGP}$) is based on the
assumption that outputs share commonalities, however, if this assumption does
not hold negative transfer will lead to decreased performance relative to
learning outputs independently or in subsets. In this article, we first define
negative transfer in the context of an $\mathcal{MGP}$ and then derive
necessary conditions for an $\mathcal{MGP}$ model to avoid negative transfer.
Specifically, under the convolution construction, we show that avoiding
negative transfer is mainly dependent on having a sufficient number of latent
functions $Q$ regardless of the flexibility of the kernel or inference
procedure used. However, a slight increase in $Q$ leads to a large increase in
the number of parameters to be estimated. To this end, we propose two latent
structures that scale to arbitrarily large datasets, can avoid negative
transfer and allow any kernel or sparse approximations to be used within. These
structures also allow regularization which can provide consistent and automatic
selection of related outputs.
| [
{
"version": "v1",
"created": "Mon, 6 Apr 2020 02:47:30 GMT"
}
] | 2023-07-04T00:00:00 | [
[
"Li",
"Moyan",
""
],
[
"Kontar",
"Raed",
""
]
] | TITLE: On Negative Transfer and Structure of Latent Functions in Multi-output
Gaussian Processes
ABSTRACT: The multi-output Gaussian process ($\mathcal{MGP}$) is based on the
assumption that outputs share commonalities, however, if this assumption does
not hold negative transfer will lead to decreased performance relative to
learning outputs independently or in subsets. In this article, we first define
negative transfer in the context of an $\mathcal{MGP}$ and then derive
necessary conditions for an $\mathcal{MGP}$ model to avoid negative transfer.
Specifically, under the convolution construction, we show that avoiding
negative transfer is mainly dependent on having a sufficient number of latent
functions $Q$ regardless of the flexibility of the kernel or inference
procedure used. However, a slight increase in $Q$ leads to a large increase in
the number of parameters to be estimated. To this end, we propose two latent
structures that scale to arbitrarily large datasets, can avoid negative
transfer and allow any kernel or sparse approximations to be used within. These
structures also allow regularization which can provide consistent and automatic
selection of related outputs.
| no_new_dataset | 0.708988 |
2007.07703 | Evan Piermont | Evan Piermont and Peio Zuazo-Garin | Failures of Contingent Thinking | null | null | null | null | cs.AI econ.TH | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this paper, we provide a theoretical framework to analyze an agent who
misinterprets or misperceives the true decision problem she faces. We show that
a wide range of behavior observed in experimental settings manifest as failures
to perceive implications, in other words, to properly account for the logical
relationships between various payoff relevant contingencies. We present a
behavioral definition of perceived implication, thereby providing an
elicitation technique, and show that an agent's account of implication
identifies a subjective state-space that underlies her behavior. By analyzing
this state-space, we characterize distinct benchmarks of logical sophistication
that drive empirical phenomena. We disentangle static and dynamic rationality.
Thus, our framework delivers both a methodology for assessing an agent's level
of contingent thinking and a strategy for identifying her beliefs in the
absence full rationality.
| [
{
"version": "v1",
"created": "Wed, 15 Jul 2020 14:21:16 GMT"
},
{
"version": "v2",
"created": "Wed, 7 Dec 2022 11:50:10 GMT"
},
{
"version": "v3",
"created": "Mon, 3 Jul 2023 12:15:09 GMT"
}
] | 2023-07-04T00:00:00 | [
[
"Piermont",
"Evan",
""
],
[
"Zuazo-Garin",
"Peio",
""
]
] | TITLE: Failures of Contingent Thinking
ABSTRACT: In this paper, we provide a theoretical framework to analyze an agent who
misinterprets or misperceives the true decision problem she faces. We show that
a wide range of behavior observed in experimental settings manifest as failures
to perceive implications, in other words, to properly account for the logical
relationships between various payoff relevant contingencies. We present a
behavioral definition of perceived implication, thereby providing an
elicitation technique, and show that an agent's account of implication
identifies a subjective state-space that underlies her behavior. By analyzing
this state-space, we characterize distinct benchmarks of logical sophistication
that drive empirical phenomena. We disentangle static and dynamic rationality.
Thus, our framework delivers both a methodology for assessing an agent's level
of contingent thinking and a strategy for identifying her beliefs in the
absence full rationality.
| no_new_dataset | 0.711186 |
2010.03290 | Taisuke Kobayashi | Taisuke Kobayashi | Proximal Policy Optimization with Relative Pearson Divergence | 6 pages, 5 figures (accepted for ICRA2021) | null | 10.1109/ICRA48506.2021.9560856 | null | cs.LG cs.RO | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The recent remarkable progress of deep reinforcement learning (DRL) stands on
regularization of policy for stable and efficient learning. A popular method,
named proximal policy optimization (PPO), has been introduced for this purpose.
PPO clips density ratio of the latest and baseline policies with a threshold,
while its minimization target is unclear. As another problem of PPO, the
symmetric threshold is given numerically while the density ratio itself is in
asymmetric domain, thereby causing unbalanced regularization of the policy.
This paper therefore proposes a new variant of PPO by considering a
regularization problem of relative Pearson (RPE) divergence, so-called PPO-RPE.
This regularization yields the clear minimization target, which constrains the
latest policy to the baseline one. Through its analysis, the intuitive
threshold-based design consistent with the asymmetry of the threshold and the
domain of density ratio can be derived. Through four benchmark tasks, PPO-RPE
performed as well as or better than the conventional methods in terms of the
task performance by the learned policy.
| [
{
"version": "v1",
"created": "Wed, 7 Oct 2020 09:11:22 GMT"
},
{
"version": "v2",
"created": "Mon, 15 Mar 2021 07:37:35 GMT"
}
] | 2023-07-04T00:00:00 | [
[
"Kobayashi",
"Taisuke",
""
]
] | TITLE: Proximal Policy Optimization with Relative Pearson Divergence
ABSTRACT: The recent remarkable progress of deep reinforcement learning (DRL) stands on
regularization of policy for stable and efficient learning. A popular method,
named proximal policy optimization (PPO), has been introduced for this purpose.
PPO clips density ratio of the latest and baseline policies with a threshold,
while its minimization target is unclear. As another problem of PPO, the
symmetric threshold is given numerically while the density ratio itself is in
asymmetric domain, thereby causing unbalanced regularization of the policy.
This paper therefore proposes a new variant of PPO by considering a
regularization problem of relative Pearson (RPE) divergence, so-called PPO-RPE.
This regularization yields the clear minimization target, which constrains the
latest policy to the baseline one. Through its analysis, the intuitive
threshold-based design consistent with the asymmetry of the threshold and the
domain of density ratio can be derived. Through four benchmark tasks, PPO-RPE
performed as well as or better than the conventional methods in terms of the
task performance by the learned policy.
| no_new_dataset | 0.709019 |
2103.12021 | Paria Rashidinejad | Paria Rashidinejad, Banghua Zhu, Cong Ma, Jiantao Jiao, Stuart Russell | Bridging Offline Reinforcement Learning and Imitation Learning: A Tale
of Pessimism | null | Published at NeurIPS 2021 and IEEE Transactions on Information
Theory | null | null | cs.LG cs.AI math.OC math.ST stat.ML stat.TH | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Offline (or batch) reinforcement learning (RL) algorithms seek to learn an
optimal policy from a fixed dataset without active data collection. Based on
the composition of the offline dataset, two main categories of methods are
used: imitation learning which is suitable for expert datasets and vanilla
offline RL which often requires uniform coverage datasets. From a practical
standpoint, datasets often deviate from these two extremes and the exact data
composition is usually unknown a priori. To bridge this gap, we present a new
offline RL framework that smoothly interpolates between the two extremes of
data composition, hence unifying imitation learning and vanilla offline RL. The
new framework is centered around a weak version of the concentrability
coefficient that measures the deviation from the behavior policy to the expert
policy alone.
Under this new framework, we further investigate the question on algorithm
design: can one develop an algorithm that achieves a minimax optimal rate and
also adapts to unknown data composition? To address this question, we consider
a lower confidence bound (LCB) algorithm developed based on pessimism in the
face of uncertainty in offline RL. We study finite-sample properties of LCB as
well as information-theoretic limits in multi-armed bandits, contextual
bandits, and Markov decision processes (MDPs). Our analysis reveals surprising
facts about optimality rates. In particular, in all three settings, LCB
achieves a faster rate of $1/N$ for nearly-expert datasets compared to the
usual rate of $1/\sqrt{N}$ in offline RL, where $N$ is the number of samples in
the batch dataset. In the case of contextual bandits with at least two
contexts, we prove that LCB is adaptively optimal for the entire data
composition range, achieving a smooth transition from imitation learning to
offline RL. We further show that LCB is almost adaptively optimal in MDPs.
| [
{
"version": "v1",
"created": "Mon, 22 Mar 2021 17:27:08 GMT"
},
{
"version": "v2",
"created": "Mon, 3 Jul 2023 04:47:42 GMT"
}
] | 2023-07-04T00:00:00 | [
[
"Rashidinejad",
"Paria",
""
],
[
"Zhu",
"Banghua",
""
],
[
"Ma",
"Cong",
""
],
[
"Jiao",
"Jiantao",
""
],
[
"Russell",
"Stuart",
""
]
] | TITLE: Bridging Offline Reinforcement Learning and Imitation Learning: A Tale
of Pessimism
ABSTRACT: Offline (or batch) reinforcement learning (RL) algorithms seek to learn an
optimal policy from a fixed dataset without active data collection. Based on
the composition of the offline dataset, two main categories of methods are
used: imitation learning which is suitable for expert datasets and vanilla
offline RL which often requires uniform coverage datasets. From a practical
standpoint, datasets often deviate from these two extremes and the exact data
composition is usually unknown a priori. To bridge this gap, we present a new
offline RL framework that smoothly interpolates between the two extremes of
data composition, hence unifying imitation learning and vanilla offline RL. The
new framework is centered around a weak version of the concentrability
coefficient that measures the deviation from the behavior policy to the expert
policy alone.
Under this new framework, we further investigate the question on algorithm
design: can one develop an algorithm that achieves a minimax optimal rate and
also adapts to unknown data composition? To address this question, we consider
a lower confidence bound (LCB) algorithm developed based on pessimism in the
face of uncertainty in offline RL. We study finite-sample properties of LCB as
well as information-theoretic limits in multi-armed bandits, contextual
bandits, and Markov decision processes (MDPs). Our analysis reveals surprising
facts about optimality rates. In particular, in all three settings, LCB
achieves a faster rate of $1/N$ for nearly-expert datasets compared to the
usual rate of $1/\sqrt{N}$ in offline RL, where $N$ is the number of samples in
the batch dataset. In the case of contextual bandits with at least two
contexts, we prove that LCB is adaptively optimal for the entire data
composition range, achieving a smooth transition from imitation learning to
offline RL. We further show that LCB is almost adaptively optimal in MDPs.
| no_new_dataset | 0.7076 |
2107.09889 | Wenxuan Liu | Wenxuan Liu, Tianyao He, Chen Gong, Ning Zhang, Hua Yang, Junchi Yan | Fine-Grained Music Plagiarism Detection: Revealing Plagiarists through
Bipartite Graph Matching and a Comprehensive Large-Scale Dataset | null | null | null | null | cs.SD cs.MM eess.AS | http://creativecommons.org/licenses/by/4.0/ | Music plagiarism detection is gaining more and more attention due to the
popularity of music production and society's emphasis on intellectual property.
We aim to find fine-grained plagiarism in music pairs since conventional
methods are coarse-grained and cannot match real-life scenarios. Considering
that there is no sizeable dataset designed for the music plagiarism task, we
establish a large-scale simulated dataset, named Music Plagiarism Detection
Dataset (MPD-Set) under the guidance and expertise of renowned researchers from
national-level professional institutions in the field of music. MPD-Set
considers diverse music plagiarism cases found in real life from the melodic,
rhythmic, and tonal levels respectively. Further, we establish a Real-life
Dataset for evaluation, where all plagiarism pairs are real cases. To detect
the fine-grained plagiarism pairs effectively, we propose a graph-based method
called Bipatite Melody Matching Detector (BMM-Det), which formulates the
problem as a max matching problem in the bipartite graph. Experimental results
on both the simulated and Real-life Datasets demonstrate that BMM-Det
outperforms the existing plagiarism detection methods, and is robust to common
plagiarism cases like transpositions, pitch shifts, duration variance, and
melody change. Datasets and source code are open-sourced at
https://github.com/xuan301/BMMDet_MPDSet.
| [
{
"version": "v1",
"created": "Wed, 21 Jul 2021 06:04:47 GMT"
},
{
"version": "v2",
"created": "Sun, 2 Jul 2023 08:28:07 GMT"
}
] | 2023-07-04T00:00:00 | [
[
"Liu",
"Wenxuan",
""
],
[
"He",
"Tianyao",
""
],
[
"Gong",
"Chen",
""
],
[
"Zhang",
"Ning",
""
],
[
"Yang",
"Hua",
""
],
[
"Yan",
"Junchi",
""
]
] | TITLE: Fine-Grained Music Plagiarism Detection: Revealing Plagiarists through
Bipartite Graph Matching and a Comprehensive Large-Scale Dataset
ABSTRACT: Music plagiarism detection is gaining more and more attention due to the
popularity of music production and society's emphasis on intellectual property.
We aim to find fine-grained plagiarism in music pairs since conventional
methods are coarse-grained and cannot match real-life scenarios. Considering
that there is no sizeable dataset designed for the music plagiarism task, we
establish a large-scale simulated dataset, named Music Plagiarism Detection
Dataset (MPD-Set) under the guidance and expertise of renowned researchers from
national-level professional institutions in the field of music. MPD-Set
considers diverse music plagiarism cases found in real life from the melodic,
rhythmic, and tonal levels respectively. Further, we establish a Real-life
Dataset for evaluation, where all plagiarism pairs are real cases. To detect
the fine-grained plagiarism pairs effectively, we propose a graph-based method
called Bipatite Melody Matching Detector (BMM-Det), which formulates the
problem as a max matching problem in the bipartite graph. Experimental results
on both the simulated and Real-life Datasets demonstrate that BMM-Det
outperforms the existing plagiarism detection methods, and is robust to common
plagiarism cases like transpositions, pitch shifts, duration variance, and
melody change. Datasets and source code are open-sourced at
https://github.com/xuan301/BMMDet_MPDSet.
| new_dataset | 0.714777 |
2107.11298 | Giuseppe Vecchio | Giuseppe Vecchio, Simone Palazzo, Concetto Spampinato | SurfaceNet: Adversarial SVBRDF Estimation from a Single Image | null | null | 10.1109/ICCV48922.2021 | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | In this paper we present SurfaceNet, an approach for estimating
spatially-varying bidirectional reflectance distribution function (SVBRDF)
material properties from a single image. We pose the problem as an image
translation task and propose a novel patch-based generative adversarial network
(GAN) that is able to produce high-quality, high-resolution surface reflectance
maps. The employment of the GAN paradigm has a twofold objective: 1) allowing
the model to recover finer details than standard translation models; 2)
reducing the domain shift between synthetic and real data distributions in an
unsupervised way. An extensive evaluation, carried out on a public benchmark of
synthetic and real images under different illumination conditions, shows that
SurfaceNet largely outperforms existing SVBRDF reconstruction methods, both
quantitatively and qualitatively. Furthermore, SurfaceNet exhibits a remarkable
ability in generating high-quality maps from real samples without any
supervision at training time.
| [
{
"version": "v1",
"created": "Fri, 23 Jul 2021 15:18:54 GMT"
}
] | 2023-07-04T00:00:00 | [
[
"Vecchio",
"Giuseppe",
""
],
[
"Palazzo",
"Simone",
""
],
[
"Spampinato",
"Concetto",
""
]
] | TITLE: SurfaceNet: Adversarial SVBRDF Estimation from a Single Image
ABSTRACT: In this paper we present SurfaceNet, an approach for estimating
spatially-varying bidirectional reflectance distribution function (SVBRDF)
material properties from a single image. We pose the problem as an image
translation task and propose a novel patch-based generative adversarial network
(GAN) that is able to produce high-quality, high-resolution surface reflectance
maps. The employment of the GAN paradigm has a twofold objective: 1) allowing
the model to recover finer details than standard translation models; 2)
reducing the domain shift between synthetic and real data distributions in an
unsupervised way. An extensive evaluation, carried out on a public benchmark of
synthetic and real images under different illumination conditions, shows that
SurfaceNet largely outperforms existing SVBRDF reconstruction methods, both
quantitatively and qualitatively. Furthermore, SurfaceNet exhibits a remarkable
ability in generating high-quality maps from real samples without any
supervision at training time.
| no_new_dataset | 0.711074 |
2107.12824 | Hiroki Kawakami | Hiroki Kawakami, Hirohisa Watanabe, Keisuke Sugiura, Hiroki Matsutani | A Low-Cost Neural ODE with Depthwise Separable Convolution for Edge
Domain Adaptation on FPGAs | null | IEICE Trans on Information and Systems (2023) | 10.1587/transinf.2022EDP7149 | null | cs.LG cs.AR | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | High-performance deep neural network (DNN)-based systems are in high demand
in edge environments. Due to its high computational complexity, it is
challenging to deploy DNNs on edge devices with strict limitations on
computational resources. In this paper, we derive a compact while
highly-accurate DNN model, termed dsODENet, by combining recently-proposed
parameter reduction techniques: Neural ODE (Ordinary Differential Equation) and
DSC (Depthwise Separable Convolution). Neural ODE exploits a similarity between
ResNet and ODE, and shares most of weight parameters among multiple layers,
which greatly reduces the memory consumption. We apply dsODENet to a domain
adaptation as a practical use case with image classification datasets. We also
propose a resource-efficient FPGA-based design for dsODENet, where all the
parameters and feature maps except for pre- and post-processing layers can be
mapped onto on-chip memories. It is implemented on Xilinx ZCU104 board and
evaluated in terms of domain adaptation accuracy, inference speed, FPGA
resource utilization, and speedup rate compared to a software counterpart. The
results demonstrate that dsODENet achieves comparable or slightly better domain
adaptation accuracy compared to our baseline Neural ODE implementation, while
the total parameter size without pre- and post-processing layers is reduced by
54.2% to 79.8%. Our FPGA implementation accelerates the inference speed by 23.8
times.
| [
{
"version": "v1",
"created": "Tue, 27 Jul 2021 13:44:13 GMT"
},
{
"version": "v2",
"created": "Tue, 1 Feb 2022 17:46:12 GMT"
},
{
"version": "v3",
"created": "Wed, 17 Aug 2022 05:37:45 GMT"
},
{
"version": "v4",
"created": "Fri, 17 Mar 2023 10:14:57 GMT"
}
] | 2023-07-04T00:00:00 | [
[
"Kawakami",
"Hiroki",
""
],
[
"Watanabe",
"Hirohisa",
""
],
[
"Sugiura",
"Keisuke",
""
],
[
"Matsutani",
"Hiroki",
""
]
] | TITLE: A Low-Cost Neural ODE with Depthwise Separable Convolution for Edge
Domain Adaptation on FPGAs
ABSTRACT: High-performance deep neural network (DNN)-based systems are in high demand
in edge environments. Due to its high computational complexity, it is
challenging to deploy DNNs on edge devices with strict limitations on
computational resources. In this paper, we derive a compact while
highly-accurate DNN model, termed dsODENet, by combining recently-proposed
parameter reduction techniques: Neural ODE (Ordinary Differential Equation) and
DSC (Depthwise Separable Convolution). Neural ODE exploits a similarity between
ResNet and ODE, and shares most of weight parameters among multiple layers,
which greatly reduces the memory consumption. We apply dsODENet to a domain
adaptation as a practical use case with image classification datasets. We also
propose a resource-efficient FPGA-based design for dsODENet, where all the
parameters and feature maps except for pre- and post-processing layers can be
mapped onto on-chip memories. It is implemented on Xilinx ZCU104 board and
evaluated in terms of domain adaptation accuracy, inference speed, FPGA
resource utilization, and speedup rate compared to a software counterpart. The
results demonstrate that dsODENet achieves comparable or slightly better domain
adaptation accuracy compared to our baseline Neural ODE implementation, while
the total parameter size without pre- and post-processing layers is reduced by
54.2% to 79.8%. Our FPGA implementation accelerates the inference speed by 23.8
times.
| no_new_dataset | 0.709007 |
2108.08988 | Tunazzina Islam | Tunazzina Islam, Dan Goldwasser | Twitter User Representation Using Weakly Supervised Graph Embedding | accepted at 16th International AAAI Conference on Web and Social
Media (ICWSM-2022), direct accept from May 2021 submission, 12 pages, minor
change for camera-ready | null | 10.1609/icwsm.v16i1.19298 | null | cs.CL cs.AI cs.CY cs.LG cs.SI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Social media platforms provide convenient means for users to participate in
multiple online activities on various contents and create fast widespread
interactions. However, this rapidly growing access has also increased the
diverse information, and characterizing user types to understand people's
lifestyle decisions shared in social media is challenging. In this paper, we
propose a weakly supervised graph embedding based framework for understanding
user types. We evaluate the user embedding learned using weak supervision over
well-being related tweets from Twitter, focusing on 'Yoga', 'Keto diet'.
Experiments on real-world datasets demonstrate that the proposed framework
outperforms the baselines for detecting user types. Finally, we illustrate data
analysis on different types of users (e.g., practitioner vs. promotional) from
our dataset. While we focus on lifestyle-related tweets (i.e., yoga, keto), our
method for constructing user representation readily generalizes to other
domains.
| [
{
"version": "v1",
"created": "Fri, 20 Aug 2021 03:54:29 GMT"
},
{
"version": "v2",
"created": "Thu, 26 Aug 2021 21:38:48 GMT"
},
{
"version": "v3",
"created": "Wed, 13 Apr 2022 23:22:02 GMT"
}
] | 2023-07-04T00:00:00 | [
[
"Islam",
"Tunazzina",
""
],
[
"Goldwasser",
"Dan",
""
]
] | TITLE: Twitter User Representation Using Weakly Supervised Graph Embedding
ABSTRACT: Social media platforms provide convenient means for users to participate in
multiple online activities on various contents and create fast widespread
interactions. However, this rapidly growing access has also increased the
diverse information, and characterizing user types to understand people's
lifestyle decisions shared in social media is challenging. In this paper, we
propose a weakly supervised graph embedding based framework for understanding
user types. We evaluate the user embedding learned using weak supervision over
well-being related tweets from Twitter, focusing on 'Yoga', 'Keto diet'.
Experiments on real-world datasets demonstrate that the proposed framework
outperforms the baselines for detecting user types. Finally, we illustrate data
analysis on different types of users (e.g., practitioner vs. promotional) from
our dataset. While we focus on lifestyle-related tweets (i.e., yoga, keto), our
method for constructing user representation readily generalizes to other
domains.
| no_new_dataset | 0.6329 |
2108.09388 | Bayan Al-Nahhas | Bayan Al-Nahhas, Qurrat-Ul-Ain Nadeem, and Anas Chaaban | Distributed Reconfigurable Intelligent Surfaces Assisted Wireless
Communication: Asymptotic Analysis under Imperfect CSI | null | null | null | null | cs.IT eess.SP math.IT | http://creativecommons.org/licenses/by/4.0/ | This work studies the net sum-rate performance of a distributed
reconfigurable intelligent surfaces (RISs)-assisted multi-user
multiple-input-single-output (MISO) downlink communication system under
imperfect instantaneous-channel state information (I-CSI) to implement
precoding at the base station (BS) and statistical-CSI (S-CSI) to design the
RISs phase-shifts. Two channel estimation (CE) protocols are considered for
I-CSI acquisition: (i) a full CE protocol that estimates all direct and
RISs-assisted channels over multiple training sub-phases, and (ii) a
low-overhead direct estimation (DE) protocol that estimates the end-to-end
channel in a single sub-phase. We derive the deterministic equivalents of
signal-to-interference-plus-noise ratio (SINR) and ergodic net sum-rate under
Rayleigh and Rician fading and both CE protocols, for given RISs phase-shifts,
which are then optimized based on S-CSI. Simulation results reveal that the
low-complexity DE protocol yields better net sum-rate than the full CE protocol
when used to obtain CSI for precoding. A benchmark full I-CSI based RISs design
is also outlined and shown to yield higher SINR but lower net sum-rate than the
S-CSI based RISs design due to the large overhead associated with full I-CSI
acquisition. Therefore the proposed DE-S-CSI based design for precoding and
reflect beamforming achieves high net sum-rate with low complexity, overhead
and power consumption.
| [
{
"version": "v1",
"created": "Fri, 20 Aug 2021 22:19:56 GMT"
},
{
"version": "v2",
"created": "Fri, 30 Jun 2023 21:16:33 GMT"
}
] | 2023-07-04T00:00:00 | [
[
"Al-Nahhas",
"Bayan",
""
],
[
"Nadeem",
"Qurrat-Ul-Ain",
""
],
[
"Chaaban",
"Anas",
""
]
] | TITLE: Distributed Reconfigurable Intelligent Surfaces Assisted Wireless
Communication: Asymptotic Analysis under Imperfect CSI
ABSTRACT: This work studies the net sum-rate performance of a distributed
reconfigurable intelligent surfaces (RISs)-assisted multi-user
multiple-input-single-output (MISO) downlink communication system under
imperfect instantaneous-channel state information (I-CSI) to implement
precoding at the base station (BS) and statistical-CSI (S-CSI) to design the
RISs phase-shifts. Two channel estimation (CE) protocols are considered for
I-CSI acquisition: (i) a full CE protocol that estimates all direct and
RISs-assisted channels over multiple training sub-phases, and (ii) a
low-overhead direct estimation (DE) protocol that estimates the end-to-end
channel in a single sub-phase. We derive the deterministic equivalents of
signal-to-interference-plus-noise ratio (SINR) and ergodic net sum-rate under
Rayleigh and Rician fading and both CE protocols, for given RISs phase-shifts,
which are then optimized based on S-CSI. Simulation results reveal that the
low-complexity DE protocol yields better net sum-rate than the full CE protocol
when used to obtain CSI for precoding. A benchmark full I-CSI based RISs design
is also outlined and shown to yield higher SINR but lower net sum-rate than the
S-CSI based RISs design due to the large overhead associated with full I-CSI
acquisition. Therefore the proposed DE-S-CSI based design for precoding and
reflect beamforming achieves high net sum-rate with low complexity, overhead
and power consumption.
| no_new_dataset | 0.710867 |
2109.04470 | Adithya Kulkarni | Nasim Sabetpour, Adithya Kulkarni, Sihong Xie, Qi Li | Truth Discovery in Sequence Labels from Crowds | null | null | null | null | cs.HC cs.LG | http://creativecommons.org/licenses/by/4.0/ | Annotation quality and quantity positively affect the learning performance of
sequence labeling, a vital task in Natural Language Processing. Hiring domain
experts to annotate a corpus is very costly in terms of money and time.
Crowdsourcing platforms, such as Amazon Mechanical Turk (AMT), have been
deployed to assist in this purpose. However, the annotations collected this way
are prone to human errors due to the lack of expertise of the crowd workers.
Existing literature in annotation aggregation assumes that annotations are
independent and thus faces challenges when handling the sequential label
aggregation tasks with complex dependencies. To conquer the challenges, we
propose an optimization-based method that infers the ground truth labels using
annotations provided by workers for sequential labeling tasks. The proposed
Aggregation method for Sequential Labels from Crowds ($AggSLC$) jointly
considers the characteristics of sequential labeling tasks, workers'
reliabilities, and advanced machine learning techniques. Theoretical analysis
on the algorithm's convergence further demonstrates that the proposed $AggSLC$
halts after a finite number of iterations. We evaluate $AggSLC$ on different
crowdsourced datasets for Named Entity Recognition (NER) tasks and Information
Extraction tasks in biomedical (PICO), as well as a simulated dataset. Our
results show that the proposed method outperforms the state-of-the-art
aggregation methods. To achieve insights into the framework, we study the
effectiveness of $AggSLC$'s components through ablation studies.
| [
{
"version": "v1",
"created": "Thu, 9 Sep 2021 19:12:13 GMT"
},
{
"version": "v2",
"created": "Sat, 1 Jul 2023 23:38:34 GMT"
}
] | 2023-07-04T00:00:00 | [
[
"Sabetpour",
"Nasim",
""
],
[
"Kulkarni",
"Adithya",
""
],
[
"Xie",
"Sihong",
""
],
[
"Li",
"Qi",
""
]
] | TITLE: Truth Discovery in Sequence Labels from Crowds
ABSTRACT: Annotation quality and quantity positively affect the learning performance of
sequence labeling, a vital task in Natural Language Processing. Hiring domain
experts to annotate a corpus is very costly in terms of money and time.
Crowdsourcing platforms, such as Amazon Mechanical Turk (AMT), have been
deployed to assist in this purpose. However, the annotations collected this way
are prone to human errors due to the lack of expertise of the crowd workers.
Existing literature in annotation aggregation assumes that annotations are
independent and thus faces challenges when handling the sequential label
aggregation tasks with complex dependencies. To conquer the challenges, we
propose an optimization-based method that infers the ground truth labels using
annotations provided by workers for sequential labeling tasks. The proposed
Aggregation method for Sequential Labels from Crowds ($AggSLC$) jointly
considers the characteristics of sequential labeling tasks, workers'
reliabilities, and advanced machine learning techniques. Theoretical analysis
on the algorithm's convergence further demonstrates that the proposed $AggSLC$
halts after a finite number of iterations. We evaluate $AggSLC$ on different
crowdsourced datasets for Named Entity Recognition (NER) tasks and Information
Extraction tasks in biomedical (PICO), as well as a simulated dataset. Our
results show that the proposed method outperforms the state-of-the-art
aggregation methods. To achieve insights into the framework, we study the
effectiveness of $AggSLC$'s components through ablation studies.
| no_new_dataset | 0.707815 |
2109.09927 | Rajendra Nagar | Rajendra Nagar | Robust Extrinsic Symmetry Estimation in 3D Point Clouds | null | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | Detecting the reflection symmetry plane of an object represented by a 3D
point cloud is a fundamental problem in 3D computer vision and geometry
processing due to its various applications, such as compression, object
detection, robotic grasping, 3D surface reconstruction, etc. There exist
several efficient approaches for solving this problem for clean 3D point
clouds. However, it is a challenging problem to solve in the presence of
outliers and missing parts. The existing methods try to overcome this challenge
mostly by voting-based techniques but do not work efficiently. In this work, we
proposed a statistical estimator-based approach for the plane of reflection
symmetry that is robust to outliers and missing parts. We pose the problem of
finding the optimal estimator for the reflection symmetry as an optimization
problem on a 2-Sphere that quickly converges to the global solution for an
approximate initialization. We further adapt the heat kernel signature for
symmetry invariant matching of mirror symmetric points. This approach helps us
to decouple the chicken-and-egg problem of finding the optimal symmetry plane
and correspondences between the reflective symmetric points. The proposed
approach achieves comparable mean ground-truth error and 4.5\% increment in the
F-score as compared to the state-of-the-art approaches on the benchmark
dataset.
| [
{
"version": "v1",
"created": "Tue, 21 Sep 2021 03:09:51 GMT"
},
{
"version": "v2",
"created": "Sun, 2 Jul 2023 06:25:17 GMT"
}
] | 2023-07-04T00:00:00 | [
[
"Nagar",
"Rajendra",
""
]
] | TITLE: Robust Extrinsic Symmetry Estimation in 3D Point Clouds
ABSTRACT: Detecting the reflection symmetry plane of an object represented by a 3D
point cloud is a fundamental problem in 3D computer vision and geometry
processing due to its various applications, such as compression, object
detection, robotic grasping, 3D surface reconstruction, etc. There exist
several efficient approaches for solving this problem for clean 3D point
clouds. However, it is a challenging problem to solve in the presence of
outliers and missing parts. The existing methods try to overcome this challenge
mostly by voting-based techniques but do not work efficiently. In this work, we
proposed a statistical estimator-based approach for the plane of reflection
symmetry that is robust to outliers and missing parts. We pose the problem of
finding the optimal estimator for the reflection symmetry as an optimization
problem on a 2-Sphere that quickly converges to the global solution for an
approximate initialization. We further adapt the heat kernel signature for
symmetry invariant matching of mirror symmetric points. This approach helps us
to decouple the chicken-and-egg problem of finding the optimal symmetry plane
and correspondences between the reflective symmetric points. The proposed
approach achieves comparable mean ground-truth error and 4.5\% increment in the
F-score as compared to the state-of-the-art approaches on the benchmark
dataset.
| no_new_dataset | 0.711638 |
2110.05674 | Liang Wang | Liang Wang, Luis Carvalho | Deviance Matrix Factorization | null | null | null | null | stat.ML cs.LG stat.AP stat.CO stat.ME | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We investigate a general matrix factorization for deviance-based data losses,
extending the ubiquitous singular value decomposition beyond squared error
loss. While similar approaches have been explored before, our method leverages
classical statistical methodology from generalized linear models (GLMs) and
provides an efficient algorithm that is flexible enough to allow for structural
zeros via entry weights. Moreover, by adapting results from GLM theory, we
provide support for these decompositions by (i) showing strong consistency
under the GLM setup, (ii) checking the adequacy of a chosen exponential family
via a generalized Hosmer-Lemeshow test, and (iii) determining the rank of the
decomposition via a maximum eigenvalue gap method. To further support our
findings, we conduct simulation studies to assess robustness to decomposition
assumptions and extensive case studies using benchmark datasets from image face
recognition, natural language processing, network analysis, and biomedical
studies. Our theoretical and empirical results indicate that the proposed
decomposition is more flexible, general, and robust, and can thus provide
improved performance when compared to similar methods. To facilitate
applications, an R package with efficient model fitting and family and rank
determination is also provided.
| [
{
"version": "v1",
"created": "Tue, 12 Oct 2021 01:27:55 GMT"
},
{
"version": "v2",
"created": "Fri, 2 Dec 2022 22:11:40 GMT"
},
{
"version": "v3",
"created": "Fri, 30 Jun 2023 23:08:23 GMT"
}
] | 2023-07-04T00:00:00 | [
[
"Wang",
"Liang",
""
],
[
"Carvalho",
"Luis",
""
]
] | TITLE: Deviance Matrix Factorization
ABSTRACT: We investigate a general matrix factorization for deviance-based data losses,
extending the ubiquitous singular value decomposition beyond squared error
loss. While similar approaches have been explored before, our method leverages
classical statistical methodology from generalized linear models (GLMs) and
provides an efficient algorithm that is flexible enough to allow for structural
zeros via entry weights. Moreover, by adapting results from GLM theory, we
provide support for these decompositions by (i) showing strong consistency
under the GLM setup, (ii) checking the adequacy of a chosen exponential family
via a generalized Hosmer-Lemeshow test, and (iii) determining the rank of the
decomposition via a maximum eigenvalue gap method. To further support our
findings, we conduct simulation studies to assess robustness to decomposition
assumptions and extensive case studies using benchmark datasets from image face
recognition, natural language processing, network analysis, and biomedical
studies. Our theoretical and empirical results indicate that the proposed
decomposition is more flexible, general, and robust, and can thus provide
improved performance when compared to similar methods. To facilitate
applications, an R package with efficient model fitting and family and rank
determination is also provided.
| no_new_dataset | 0.707771 |
2111.03892 | Bo Lyu | Bo Lyu, Shiping Wen | TND-NAS: Towards Non-differentiable Objectives in Progressive
Differentiable NAS Framework | null | null | null | null | cs.LG cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Differentiable architecture search has gradually become the mainstream
research topic in the field of Neural Architecture Search (NAS) for its high
efficiency compared with the early NAS methods. Recent differentiable NAS also
aims at further improving the search performance and reducing the GPU-memory
consumption. However, these methods are no longer naturally capable of tackling
the non-differentiable objectives, e.g., energy, resource-constrained
efficiency, and other metrics, let alone the multi-objective search demands.
Researches in the multi-objective NAS field target this but requires vast
computational resources cause of the sole optimization of each candidate
architecture. In light of this discrepancy, we propose the TND-NAS, which is
with the merits of the high efficiency in differentiable NAS framework and the
compatibility among non-differentiable metrics in Multi-objective NAS. Under
the differentiable NAS framework, with the continuous relaxation of the search
space, TND-NAS has the architecture parameters been optimized in discrete
space, while resorting to the progressive search space shrinking by
architecture parameters. Our representative experiment takes two objectives
(Parameters, Accuracy) as an example, we achieve a series of high-performance
compact architectures on CIFAR10 (1.09M/3.3%, 2.4M/2.95%, 9.57M/2.54%) and
CIFAR100 (2.46M/18.3%, 5.46/16.73%, 12.88/15.20%) datasets. Favorably, compared
with other multi-objective NAS methods, TND-NAS is less time-consuming (1.3
GPU-days on NVIDIA 1080Ti, 1/6 of that in NSGA-Net), and can be conveniently
adapted to real-world NAS scenarios (resource-constrained,
platform-specialized).
| [
{
"version": "v1",
"created": "Sat, 6 Nov 2021 14:19:36 GMT"
},
{
"version": "v2",
"created": "Thu, 21 Apr 2022 03:32:26 GMT"
},
{
"version": "v3",
"created": "Mon, 12 Jun 2023 12:39:37 GMT"
},
{
"version": "v4",
"created": "Sat, 1 Jul 2023 12:36:56 GMT"
}
] | 2023-07-04T00:00:00 | [
[
"Lyu",
"Bo",
""
],
[
"Wen",
"Shiping",
""
]
] | TITLE: TND-NAS: Towards Non-differentiable Objectives in Progressive
Differentiable NAS Framework
ABSTRACT: Differentiable architecture search has gradually become the mainstream
research topic in the field of Neural Architecture Search (NAS) for its high
efficiency compared with the early NAS methods. Recent differentiable NAS also
aims at further improving the search performance and reducing the GPU-memory
consumption. However, these methods are no longer naturally capable of tackling
the non-differentiable objectives, e.g., energy, resource-constrained
efficiency, and other metrics, let alone the multi-objective search demands.
Researches in the multi-objective NAS field target this but requires vast
computational resources cause of the sole optimization of each candidate
architecture. In light of this discrepancy, we propose the TND-NAS, which is
with the merits of the high efficiency in differentiable NAS framework and the
compatibility among non-differentiable metrics in Multi-objective NAS. Under
the differentiable NAS framework, with the continuous relaxation of the search
space, TND-NAS has the architecture parameters been optimized in discrete
space, while resorting to the progressive search space shrinking by
architecture parameters. Our representative experiment takes two objectives
(Parameters, Accuracy) as an example, we achieve a series of high-performance
compact architectures on CIFAR10 (1.09M/3.3%, 2.4M/2.95%, 9.57M/2.54%) and
CIFAR100 (2.46M/18.3%, 5.46/16.73%, 12.88/15.20%) datasets. Favorably, compared
with other multi-objective NAS methods, TND-NAS is less time-consuming (1.3
GPU-days on NVIDIA 1080Ti, 1/6 of that in NSGA-Net), and can be conveniently
adapted to real-world NAS scenarios (resource-constrained,
platform-specialized).
| no_new_dataset | 0.710239 |
2111.10461 | Raed Al Kontar | Hao Chen, Lili Zheng, Raed Al Kontar, Garvesh Raskutti | Gaussian Process Inference Using Mini-batch Stochastic Gradient Descent:
Convergence Guarantees and Empirical Benefits | null | Journal of Machine learning Research (JMLR), 2022 | null | 23(227):1-59 | stat.ML cs.LG | http://creativecommons.org/licenses/by/4.0/ | Stochastic gradient descent (SGD) and its variants have established
themselves as the go-to algorithms for large-scale machine learning problems
with independent samples due to their generalization performance and intrinsic
computational advantage. However, the fact that the stochastic gradient is a
biased estimator of the full gradient with correlated samples has led to the
lack of theoretical understanding of how SGD behaves under correlated settings
and hindered its use in such cases. In this paper, we focus on hyperparameter
estimation for the Gaussian process (GP) and take a step forward towards
breaking the barrier by proving minibatch SGD converges to a critical point of
the full log-likelihood loss function, and recovers model hyperparameters with
rate $O(\frac{1}{K})$ for $K$ iterations, up to a statistical error term
depending on the minibatch size. Our theoretical guarantees hold provided that
the kernel functions exhibit exponential or polynomial eigendecay which is
satisfied by a wide range of kernels commonly used in GPs. Numerical studies on
both simulated and real datasets demonstrate that minibatch SGD has better
generalization over state-of-the-art GP methods while reducing the
computational burden and opening a new, previously unexplored, data size regime
for GPs.
| [
{
"version": "v1",
"created": "Fri, 19 Nov 2021 22:28:47 GMT"
}
] | 2023-07-04T00:00:00 | [
[
"Chen",
"Hao",
""
],
[
"Zheng",
"Lili",
""
],
[
"Kontar",
"Raed Al",
""
],
[
"Raskutti",
"Garvesh",
""
]
] | TITLE: Gaussian Process Inference Using Mini-batch Stochastic Gradient Descent:
Convergence Guarantees and Empirical Benefits
ABSTRACT: Stochastic gradient descent (SGD) and its variants have established
themselves as the go-to algorithms for large-scale machine learning problems
with independent samples due to their generalization performance and intrinsic
computational advantage. However, the fact that the stochastic gradient is a
biased estimator of the full gradient with correlated samples has led to the
lack of theoretical understanding of how SGD behaves under correlated settings
and hindered its use in such cases. In this paper, we focus on hyperparameter
estimation for the Gaussian process (GP) and take a step forward towards
breaking the barrier by proving minibatch SGD converges to a critical point of
the full log-likelihood loss function, and recovers model hyperparameters with
rate $O(\frac{1}{K})$ for $K$ iterations, up to a statistical error term
depending on the minibatch size. Our theoretical guarantees hold provided that
the kernel functions exhibit exponential or polynomial eigendecay which is
satisfied by a wide range of kernels commonly used in GPs. Numerical studies on
both simulated and real datasets demonstrate that minibatch SGD has better
generalization over state-of-the-art GP methods while reducing the
computational burden and opening a new, previously unexplored, data size regime
for GPs.
| no_new_dataset | 0.708603 |
2112.03549 | Binglu Wang | Binglu Wang, Tao Hu, Baoshan Li, Xiaojuan Chen, Zhijie Zhang | GaTector: A Unified Framework for Gaze Object Prediction | CVPR 2022, camera ready | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Gaze object prediction is a newly proposed task that aims to discover the
objects being stared at by humans. It is of great application significance but
still lacks a unified solution framework. An intuitive solution is to
incorporate an object detection branch into an existing gaze prediction method.
However, previous gaze prediction methods usually use two different networks to
extract features from scene image and head image, which would lead to heavy
network architecture and prevent each branch from joint optimization. In this
paper, we build a novel framework named GaTector to tackle the gaze object
prediction problem in a unified way. Particularly, a specific-general-specific
(SGS) feature extractor is firstly proposed to utilize a shared backbone to
extract general features for both scene and head images. To better consider the
specificity of inputs and tasks, SGS introduces two input-specific blocks
before the shared backbone and three task-specific blocks after the shared
backbone. Specifically, a novel Defocus layer is designed to generate
object-specific features for the object detection task without losing
information or requiring extra computations. Moreover, the energy aggregation
loss is introduced to guide the gaze heatmap to concentrate on the stared box.
In the end, we propose a novel wUoC metric that can reveal the difference
between boxes even when they share no overlapping area. Extensive experiments
on the GOO dataset verify the superiority of our method in all three tracks,
i.e. object detection, gaze estimation, and gaze object prediction.
| [
{
"version": "v1",
"created": "Tue, 7 Dec 2021 07:50:03 GMT"
},
{
"version": "v2",
"created": "Sun, 20 Mar 2022 09:48:39 GMT"
},
{
"version": "v3",
"created": "Sat, 1 Jul 2023 02:29:33 GMT"
}
] | 2023-07-04T00:00:00 | [
[
"Wang",
"Binglu",
""
],
[
"Hu",
"Tao",
""
],
[
"Li",
"Baoshan",
""
],
[
"Chen",
"Xiaojuan",
""
],
[
"Zhang",
"Zhijie",
""
]
] | TITLE: GaTector: A Unified Framework for Gaze Object Prediction
ABSTRACT: Gaze object prediction is a newly proposed task that aims to discover the
objects being stared at by humans. It is of great application significance but
still lacks a unified solution framework. An intuitive solution is to
incorporate an object detection branch into an existing gaze prediction method.
However, previous gaze prediction methods usually use two different networks to
extract features from scene image and head image, which would lead to heavy
network architecture and prevent each branch from joint optimization. In this
paper, we build a novel framework named GaTector to tackle the gaze object
prediction problem in a unified way. Particularly, a specific-general-specific
(SGS) feature extractor is firstly proposed to utilize a shared backbone to
extract general features for both scene and head images. To better consider the
specificity of inputs and tasks, SGS introduces two input-specific blocks
before the shared backbone and three task-specific blocks after the shared
backbone. Specifically, a novel Defocus layer is designed to generate
object-specific features for the object detection task without losing
information or requiring extra computations. Moreover, the energy aggregation
loss is introduced to guide the gaze heatmap to concentrate on the stared box.
In the end, we propose a novel wUoC metric that can reveal the difference
between boxes even when they share no overlapping area. Extensive experiments
on the GOO dataset verify the superiority of our method in all three tracks,
i.e. object detection, gaze estimation, and gaze object prediction.
| no_new_dataset | 0.707645 |
2112.14900 | Zijian Li | Xuexin Chen, Ruichu Cai, Yuan Fang, Min Wu, Zijian Li, Zhifeng Hao | Motif Graph Neural Network | null | null | null | null | cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Graphs can model complicated interactions between entities, which naturally
emerge in many important applications. These applications can often be cast
into standard graph learning tasks, in which a crucial step is to learn
low-dimensional graph representations. Graph neural networks (GNNs) are
currently the most popular model in graph embedding approaches. However,
standard GNNs in the neighborhood aggregation paradigm suffer from limited
discriminative power in distinguishing \emph{high-order} graph structures as
opposed to \emph{low-order} structures. To capture high-order structures,
researchers have resorted to motifs and developed motif-based GNNs. However,
existing motif-based GNNs still often suffer from less discriminative power on
high-order structures. To overcome the above limitations, we propose Motif
Graph Neural Network (MGNN), a novel framework to better capture high-order
structures, hinging on our proposed motif redundancy minimization operator and
injective motif combination. First, MGNN produces a set of node representations
w.r.t. each motif. The next phase is our proposed redundancy minimization among
motifs which compares the motifs with each other and distills the features
unique to each motif. Finally, MGNN performs the updating of node
representations by combining multiple representations from different motifs. In
particular, to enhance the discriminative power, MGNN utilizes an injective
function to combine the representations w.r.t. different motifs. We further
show that our proposed architecture increases the expressive power of GNNs with
a theoretical analysis. We demonstrate that MGNN outperforms state-of-the-art
methods on seven public benchmarks on both node classification and graph
classification tasks.
| [
{
"version": "v1",
"created": "Thu, 30 Dec 2021 02:57:36 GMT"
},
{
"version": "v2",
"created": "Mon, 17 Jan 2022 09:24:25 GMT"
},
{
"version": "v3",
"created": "Mon, 3 Jul 2023 09:27:25 GMT"
}
] | 2023-07-04T00:00:00 | [
[
"Chen",
"Xuexin",
""
],
[
"Cai",
"Ruichu",
""
],
[
"Fang",
"Yuan",
""
],
[
"Wu",
"Min",
""
],
[
"Li",
"Zijian",
""
],
[
"Hao",
"Zhifeng",
""
]
] | TITLE: Motif Graph Neural Network
ABSTRACT: Graphs can model complicated interactions between entities, which naturally
emerge in many important applications. These applications can often be cast
into standard graph learning tasks, in which a crucial step is to learn
low-dimensional graph representations. Graph neural networks (GNNs) are
currently the most popular model in graph embedding approaches. However,
standard GNNs in the neighborhood aggregation paradigm suffer from limited
discriminative power in distinguishing \emph{high-order} graph structures as
opposed to \emph{low-order} structures. To capture high-order structures,
researchers have resorted to motifs and developed motif-based GNNs. However,
existing motif-based GNNs still often suffer from less discriminative power on
high-order structures. To overcome the above limitations, we propose Motif
Graph Neural Network (MGNN), a novel framework to better capture high-order
structures, hinging on our proposed motif redundancy minimization operator and
injective motif combination. First, MGNN produces a set of node representations
w.r.t. each motif. The next phase is our proposed redundancy minimization among
motifs which compares the motifs with each other and distills the features
unique to each motif. Finally, MGNN performs the updating of node
representations by combining multiple representations from different motifs. In
particular, to enhance the discriminative power, MGNN utilizes an injective
function to combine the representations w.r.t. different motifs. We further
show that our proposed architecture increases the expressive power of GNNs with
a theoretical analysis. We demonstrate that MGNN outperforms state-of-the-art
methods on seven public benchmarks on both node classification and graph
classification tasks.
| no_new_dataset | 0.711049 |
2201.05768 | Lishun Wang | Lishun Wang, Zongliang Wu, Yong Zhong, Xin Yuan | Spectral Compressive Imaging Reconstruction Using Convolution and
Contextual Transformer | null | null | null | null | eess.IV cs.CV | http://creativecommons.org/licenses/by/4.0/ | Spectral compressive imaging (SCI) is able to encode the high-dimensional
hyperspectral image to a 2D measurement, and then uses algorithms to
reconstruct the spatio-spectral data-cube. At present, the main bottleneck of
SCI is the reconstruction algorithm, and the state-of-the-art (SOTA)
reconstruction methods generally face the problem of long reconstruction time
and/or poor detail recovery. In this paper, we propose a novel hybrid network
module, namely CCoT (Convolution and Contextual Transformer) block, which can
acquire the inductive bias ability of convolution and the powerful modeling
ability of transformer simultaneously,and is conducive to improving the quality
of reconstruction to restore fine details. We integrate the proposed CCoT block
into deep unfolding framework based on the generalized alternating projection
algorithm, and further propose the GAP-CCoT network. Through the experiments of
extensive synthetic and real data, our proposed model achieves higher
reconstruction quality ($>$2dB in PSNR on simulated benchmark datasets) and
shorter running time than existing SOTA algorithms by a large margin. The code
and models are publicly available at https://github.com/ucaswangls/GAP-CCoT.
| [
{
"version": "v1",
"created": "Sat, 15 Jan 2022 06:30:03 GMT"
},
{
"version": "v2",
"created": "Wed, 9 Mar 2022 13:52:27 GMT"
},
{
"version": "v3",
"created": "Thu, 28 Jul 2022 09:26:13 GMT"
},
{
"version": "v4",
"created": "Sun, 2 Jul 2023 06:44:05 GMT"
}
] | 2023-07-04T00:00:00 | [
[
"Wang",
"Lishun",
""
],
[
"Wu",
"Zongliang",
""
],
[
"Zhong",
"Yong",
""
],
[
"Yuan",
"Xin",
""
]
] | TITLE: Spectral Compressive Imaging Reconstruction Using Convolution and
Contextual Transformer
ABSTRACT: Spectral compressive imaging (SCI) is able to encode the high-dimensional
hyperspectral image to a 2D measurement, and then uses algorithms to
reconstruct the spatio-spectral data-cube. At present, the main bottleneck of
SCI is the reconstruction algorithm, and the state-of-the-art (SOTA)
reconstruction methods generally face the problem of long reconstruction time
and/or poor detail recovery. In this paper, we propose a novel hybrid network
module, namely CCoT (Convolution and Contextual Transformer) block, which can
acquire the inductive bias ability of convolution and the powerful modeling
ability of transformer simultaneously,and is conducive to improving the quality
of reconstruction to restore fine details. We integrate the proposed CCoT block
into deep unfolding framework based on the generalized alternating projection
algorithm, and further propose the GAP-CCoT network. Through the experiments of
extensive synthetic and real data, our proposed model achieves higher
reconstruction quality ($>$2dB in PSNR on simulated benchmark datasets) and
shorter running time than existing SOTA algorithms by a large margin. The code
and models are publicly available at https://github.com/ucaswangls/GAP-CCoT.
| no_new_dataset | 0.709818 |
2201.07905 | Adithya Kulkarni | Adithya Kulkarni, Nasim Sabetpour, Alexey Markin, Oliver Eulenstein,
Qi Li | CPTAM: Constituency Parse Tree Aggregation Method | null | null | null | null | cs.CL cs.AI | http://creativecommons.org/licenses/by/4.0/ | Diverse Natural Language Processing tasks employ constituency parsing to
understand the syntactic structure of a sentence according to a phrase
structure grammar. Many state-of-the-art constituency parsers are proposed, but
they may provide different results for the same sentences, especially for
corpora outside their training domains. This paper adopts the truth discovery
idea to aggregate constituency parse trees from different parsers by estimating
their reliability in the absence of ground truth. Our goal is to consistently
obtain high-quality aggregated constituency parse trees. We formulate the
constituency parse tree aggregation problem in two steps, structure aggregation
and constituent label aggregation. Specifically, we propose the first truth
discovery solution for tree structures by minimizing the weighted sum of
Robinson-Foulds (RF) distances, a classic symmetric distance metric between two
trees. Extensive experiments are conducted on benchmark datasets in different
languages and domains. The experimental results show that our method, CPTAM,
outperforms the state-of-the-art aggregation baselines. We also demonstrate
that the weights estimated by CPTAM can adequately evaluate constituency
parsers in the absence of ground truth.
| [
{
"version": "v1",
"created": "Wed, 19 Jan 2022 23:05:37 GMT"
},
{
"version": "v2",
"created": "Sat, 1 Jul 2023 23:18:06 GMT"
}
] | 2023-07-04T00:00:00 | [
[
"Kulkarni",
"Adithya",
""
],
[
"Sabetpour",
"Nasim",
""
],
[
"Markin",
"Alexey",
""
],
[
"Eulenstein",
"Oliver",
""
],
[
"Li",
"Qi",
""
]
] | TITLE: CPTAM: Constituency Parse Tree Aggregation Method
ABSTRACT: Diverse Natural Language Processing tasks employ constituency parsing to
understand the syntactic structure of a sentence according to a phrase
structure grammar. Many state-of-the-art constituency parsers are proposed, but
they may provide different results for the same sentences, especially for
corpora outside their training domains. This paper adopts the truth discovery
idea to aggregate constituency parse trees from different parsers by estimating
their reliability in the absence of ground truth. Our goal is to consistently
obtain high-quality aggregated constituency parse trees. We formulate the
constituency parse tree aggregation problem in two steps, structure aggregation
and constituent label aggregation. Specifically, we propose the first truth
discovery solution for tree structures by minimizing the weighted sum of
Robinson-Foulds (RF) distances, a classic symmetric distance metric between two
trees. Extensive experiments are conducted on benchmark datasets in different
languages and domains. The experimental results show that our method, CPTAM,
outperforms the state-of-the-art aggregation baselines. We also demonstrate
that the weights estimated by CPTAM can adequately evaluate constituency
parsers in the absence of ground truth.
| no_new_dataset | 0.712076 |
2201.09057 | Fitsum Tilahun | Fitsum Debebe Tilahun, Ameha Tsegaye Abebe, and Chung G. Kang | Multi-Agent Reinforcement Learning for Distributed Joint Communication
and Computing Resource Allocation over Cell-Free Massive MIMO-enabled Mobile
Edge Computing Network | null | null | null | null | cs.NI cs.SY eess.SY | http://creativecommons.org/licenses/by-sa/4.0/ | To support the newly introduced multimedia services with ultra-low latency
and extensive computation requirements, resource-constrained end user devices
should utilize the ubiquitous computing resources available at network edge for
augmenting on-board (local) processing with edge computing. In this regard, the
capability of cell-free massive MIMO to provide reliable access links by
guaranteeing uniform quality of service without cell edge can be exploited for
seamless parallel processing. Taking this into account, we consider a cell-free
massive MIMO-enabled mobile edge network to meet the stringent requirements of
the advanced services. For the considered mobile edge network, we formulate a
joint communication and computing resource allocation (JCCRA) problem with the
objective of minimizing energy consumption of the users while meeting the tight
delay constraints. We then propose a fully distributed cooperative solution
approach based on multiagent deep deterministic policy gradient (MADDPG)
algorithm. The simulation results demonstrate that the performance of the
proposed distributed approach has converged to that of a centralized deep
deterministic policy gradient (DDPG)-based target benchmark, while alleviating
the large overhead associated with the latter. Furthermore, it has been shown
that our approach significantly outperforms heuristic baselines in terms of
energy efficiency, roughly up to 5 times less total energy consumption.
| [
{
"version": "v1",
"created": "Sat, 4 Dec 2021 03:05:31 GMT"
},
{
"version": "v2",
"created": "Thu, 19 Jan 2023 02:40:17 GMT"
},
{
"version": "v3",
"created": "Sat, 1 Jul 2023 08:44:09 GMT"
}
] | 2023-07-04T00:00:00 | [
[
"Tilahun",
"Fitsum Debebe",
""
],
[
"Abebe",
"Ameha Tsegaye",
""
],
[
"Kang",
"Chung G.",
""
]
] | TITLE: Multi-Agent Reinforcement Learning for Distributed Joint Communication
and Computing Resource Allocation over Cell-Free Massive MIMO-enabled Mobile
Edge Computing Network
ABSTRACT: To support the newly introduced multimedia services with ultra-low latency
and extensive computation requirements, resource-constrained end user devices
should utilize the ubiquitous computing resources available at network edge for
augmenting on-board (local) processing with edge computing. In this regard, the
capability of cell-free massive MIMO to provide reliable access links by
guaranteeing uniform quality of service without cell edge can be exploited for
seamless parallel processing. Taking this into account, we consider a cell-free
massive MIMO-enabled mobile edge network to meet the stringent requirements of
the advanced services. For the considered mobile edge network, we formulate a
joint communication and computing resource allocation (JCCRA) problem with the
objective of minimizing energy consumption of the users while meeting the tight
delay constraints. We then propose a fully distributed cooperative solution
approach based on multiagent deep deterministic policy gradient (MADDPG)
algorithm. The simulation results demonstrate that the performance of the
proposed distributed approach has converged to that of a centralized deep
deterministic policy gradient (DDPG)-based target benchmark, while alleviating
the large overhead associated with the latter. Furthermore, it has been shown
that our approach significantly outperforms heuristic baselines in terms of
energy efficiency, roughly up to 5 times less total energy consumption.
| no_new_dataset | 0.707998 |
2203.09809 | Taisuke Kobayashi | Taisuke Kobayashi | Proximal Policy Optimization with Adaptive Threshold for Symmetric
Relative Density Ratio | 12 pages, 9 figures | Results in Control and Optimization, 2023 | 10.1016/j.rico.2022.100192 | null | cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Deep reinforcement learning (DRL) is one of the promising approaches for
introducing robots into complicated environments. The recent remarkable
progress of DRL stands on regularization of policy, which allows the policy to
improve stably and efficiently. A popular method, so-called proximal policy
optimization (PPO), and its variants constrain density ratio of the latest and
baseline policies when the density ratio exceeds a given threshold. This
threshold can be designed relatively intuitively, and in fact its recommended
value range has been suggested. However, the density ratio is asymmetric for
its center, and the possible error scale from its center, which should be close
to the threshold, would depend on how the baseline policy is given. In order to
maximize the values of regularization of policy, this paper proposes a new PPO
derived using relative Pearson (RPE) divergence, therefore so-called PPO-RPE,
to design the threshold adaptively. In PPO-RPE, the relative density ratio,
which can be formed with symmetry, replaces the raw density ratio. Thanks to
this symmetry, its error scale from center can easily be estimated, hence, the
threshold can be adapted for the estimated error scale. From three simple
benchmark simulations, the importance of algorithm-dependent threshold design
is revealed. By simulating additional four locomotion tasks, it is verified
that the proposed method statistically contributes to task accomplishment by
appropriately restricting the policy updates.
| [
{
"version": "v1",
"created": "Fri, 18 Mar 2022 09:13:13 GMT"
}
] | 2023-07-04T00:00:00 | [
[
"Kobayashi",
"Taisuke",
""
]
] | TITLE: Proximal Policy Optimization with Adaptive Threshold for Symmetric
Relative Density Ratio
ABSTRACT: Deep reinforcement learning (DRL) is one of the promising approaches for
introducing robots into complicated environments. The recent remarkable
progress of DRL stands on regularization of policy, which allows the policy to
improve stably and efficiently. A popular method, so-called proximal policy
optimization (PPO), and its variants constrain density ratio of the latest and
baseline policies when the density ratio exceeds a given threshold. This
threshold can be designed relatively intuitively, and in fact its recommended
value range has been suggested. However, the density ratio is asymmetric for
its center, and the possible error scale from its center, which should be close
to the threshold, would depend on how the baseline policy is given. In order to
maximize the values of regularization of policy, this paper proposes a new PPO
derived using relative Pearson (RPE) divergence, therefore so-called PPO-RPE,
to design the threshold adaptively. In PPO-RPE, the relative density ratio,
which can be formed with symmetry, replaces the raw density ratio. Thanks to
this symmetry, its error scale from center can easily be estimated, hence, the
threshold can be adapted for the estimated error scale. From three simple
benchmark simulations, the importance of algorithm-dependent threshold design
is revealed. By simulating additional four locomotion tasks, it is verified
that the proposed method statistically contributes to task accomplishment by
appropriately restricting the policy updates.
| no_new_dataset | 0.708175 |
2204.02937 | Andrew Wilson | Polina Kirichenko, Pavel Izmailov, Andrew Gordon Wilson | Last Layer Re-Training is Sufficient for Robustness to Spurious
Correlations | ICLR 2023. Code is available at
https://github.com/PolinaKirichenko/deep_feature_reweighting | null | null | null | cs.LG cs.CV stat.ML | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Neural network classifiers can largely rely on simple spurious features, such
as backgrounds, to make predictions. However, even in these cases, we show that
they still often learn core features associated with the desired attributes of
the data, contrary to recent findings. Inspired by this insight, we demonstrate
that simple last layer retraining can match or outperform state-of-the-art
approaches on spurious correlation benchmarks, but with profoundly lower
complexity and computational expenses. Moreover, we show that last layer
retraining on large ImageNet-trained models can also significantly reduce
reliance on background and texture information, improving robustness to
covariate shift, after only minutes of training on a single GPU.
| [
{
"version": "v1",
"created": "Wed, 6 Apr 2022 16:55:41 GMT"
},
{
"version": "v2",
"created": "Fri, 30 Jun 2023 22:51:42 GMT"
}
] | 2023-07-04T00:00:00 | [
[
"Kirichenko",
"Polina",
""
],
[
"Izmailov",
"Pavel",
""
],
[
"Wilson",
"Andrew Gordon",
""
]
] | TITLE: Last Layer Re-Training is Sufficient for Robustness to Spurious
Correlations
ABSTRACT: Neural network classifiers can largely rely on simple spurious features, such
as backgrounds, to make predictions. However, even in these cases, we show that
they still often learn core features associated with the desired attributes of
the data, contrary to recent findings. Inspired by this insight, we demonstrate
that simple last layer retraining can match or outperform state-of-the-art
approaches on spurious correlation benchmarks, but with profoundly lower
complexity and computational expenses. Moreover, we show that last layer
retraining on large ImageNet-trained models can also significantly reduce
reliance on background and texture information, improving robustness to
covariate shift, after only minutes of training on a single GPU.
| no_new_dataset | 0.711675 |
2204.07692 | Yo-Seb Jeon | Yongjeong Oh, Yo-Seb Jeon, Mingzhe Chen, and Walid Saad | FedVQCS: Federated Learning via Vector Quantized Compressed Sensing | null | null | 10.1109/TWC.2023.3291877 | null | cs.IT cs.DC eess.SP math.IT | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this paper, a new communication-efficient federated learning (FL)
framework is proposed, inspired by vector quantized compressed sensing. The
basic strategy of the proposed framework is to compress the local model update
at each device by applying dimensionality reduction followed by vector
quantization. Subsequently, the global model update is reconstructed at a
parameter server by applying a sparse signal recovery algorithm to the
aggregation of the compressed local model updates. By harnessing the benefits
of both dimensionality reduction and vector quantization, the proposed
framework effectively reduces the communication overhead of local update
transmissions. Both the design of the vector quantizer and the key parameters
for the compression are optimized so as to minimize the reconstruction error of
the global model update under the constraint of wireless link capacity. By
considering the reconstruction error, the convergence rate of the proposed
framework is also analyzed for a non-convex loss function. Simulation results
on the MNIST and FEMNIST datasets demonstrate that the proposed framework
provides more than a 2.4% increase in classification accuracy compared to
state-of-the-art FL frameworks when the communication overhead of the local
model update transmission is 0.1 bit per local model entry.
| [
{
"version": "v1",
"created": "Sat, 16 Apr 2022 01:03:08 GMT"
},
{
"version": "v2",
"created": "Mon, 8 Aug 2022 13:03:50 GMT"
},
{
"version": "v3",
"created": "Sat, 1 Jul 2023 00:55:23 GMT"
}
] | 2023-07-04T00:00:00 | [
[
"Oh",
"Yongjeong",
""
],
[
"Jeon",
"Yo-Seb",
""
],
[
"Chen",
"Mingzhe",
""
],
[
"Saad",
"Walid",
""
]
] | TITLE: FedVQCS: Federated Learning via Vector Quantized Compressed Sensing
ABSTRACT: In this paper, a new communication-efficient federated learning (FL)
framework is proposed, inspired by vector quantized compressed sensing. The
basic strategy of the proposed framework is to compress the local model update
at each device by applying dimensionality reduction followed by vector
quantization. Subsequently, the global model update is reconstructed at a
parameter server by applying a sparse signal recovery algorithm to the
aggregation of the compressed local model updates. By harnessing the benefits
of both dimensionality reduction and vector quantization, the proposed
framework effectively reduces the communication overhead of local update
transmissions. Both the design of the vector quantizer and the key parameters
for the compression are optimized so as to minimize the reconstruction error of
the global model update under the constraint of wireless link capacity. By
considering the reconstruction error, the convergence rate of the proposed
framework is also analyzed for a non-convex loss function. Simulation results
on the MNIST and FEMNIST datasets demonstrate that the proposed framework
provides more than a 2.4% increase in classification accuracy compared to
state-of-the-art FL frameworks when the communication overhead of the local
model update transmission is 0.1 bit per local model entry.
| no_new_dataset | 0.708988 |
2204.10037 | Taoran Fang | Taoran Fang, Zhiqing Xiao, Chunping Wang, Jiarong Xu, Xuan Yang, Yang
Yang | DropMessage: Unifying Random Dropping for Graph Neural Networks | null | Proceedings of the 37th AAAI Conference on Artificial Intelligence
(2023) | 10.1609/aaai.v37i4.25545 | 37(4), 4267-4275 | cs.LG cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Graph Neural Networks (GNNs) are powerful tools for graph representation
learning. Despite their rapid development, GNNs also face some challenges, such
as over-fitting, over-smoothing, and non-robustness. Previous works indicate
that these problems can be alleviated by random dropping methods, which
integrate augmented data into models by randomly masking parts of the input.
However, some open problems of random dropping on GNNs remain to be solved.
First, it is challenging to find a universal method that are suitable for all
cases considering the divergence of different datasets and models. Second,
augmented data introduced to GNNs causes the incomplete coverage of parameters
and unstable training process. Third, there is no theoretical analysis on the
effectiveness of random dropping methods on GNNs. In this paper, we propose a
novel random dropping method called DropMessage, which performs dropping
operations directly on the propagated messages during the message-passing
process. More importantly, we find that DropMessage provides a unified
framework for most existing random dropping methods, based on which we give
theoretical analysis of their effectiveness. Furthermore, we elaborate the
superiority of DropMessage: it stabilizes the training process by reducing
sample variance; it keeps information diversity from the perspective of
information theory, enabling it become a theoretical upper bound of other
methods. To evaluate our proposed method, we conduct experiments that aims for
multiple tasks on five public datasets and two industrial datasets with various
backbone models. The experimental results show that DropMessage has the
advantages of both effectiveness and generalization, and can significantly
alleviate the problems mentioned above.
| [
{
"version": "v1",
"created": "Thu, 21 Apr 2022 11:42:49 GMT"
},
{
"version": "v2",
"created": "Wed, 30 Nov 2022 09:17:58 GMT"
},
{
"version": "v3",
"created": "Wed, 8 Feb 2023 15:09:52 GMT"
}
] | 2023-07-04T00:00:00 | [
[
"Fang",
"Taoran",
""
],
[
"Xiao",
"Zhiqing",
""
],
[
"Wang",
"Chunping",
""
],
[
"Xu",
"Jiarong",
""
],
[
"Yang",
"Xuan",
""
],
[
"Yang",
"Yang",
""
]
] | TITLE: DropMessage: Unifying Random Dropping for Graph Neural Networks
ABSTRACT: Graph Neural Networks (GNNs) are powerful tools for graph representation
learning. Despite their rapid development, GNNs also face some challenges, such
as over-fitting, over-smoothing, and non-robustness. Previous works indicate
that these problems can be alleviated by random dropping methods, which
integrate augmented data into models by randomly masking parts of the input.
However, some open problems of random dropping on GNNs remain to be solved.
First, it is challenging to find a universal method that are suitable for all
cases considering the divergence of different datasets and models. Second,
augmented data introduced to GNNs causes the incomplete coverage of parameters
and unstable training process. Third, there is no theoretical analysis on the
effectiveness of random dropping methods on GNNs. In this paper, we propose a
novel random dropping method called DropMessage, which performs dropping
operations directly on the propagated messages during the message-passing
process. More importantly, we find that DropMessage provides a unified
framework for most existing random dropping methods, based on which we give
theoretical analysis of their effectiveness. Furthermore, we elaborate the
superiority of DropMessage: it stabilizes the training process by reducing
sample variance; it keeps information diversity from the perspective of
information theory, enabling it become a theoretical upper bound of other
methods. To evaluate our proposed method, we conduct experiments that aims for
multiple tasks on five public datasets and two industrial datasets with various
backbone models. The experimental results show that DropMessage has the
advantages of both effectiveness and generalization, and can significantly
alleviate the problems mentioned above.
| no_new_dataset | 0.708175 |
2205.00395 | Shabnam Behzad | Shabnam Behzad, Keisuke Sakaguchi, Nathan Schneider, Amir Zeldes | ELQA: A Corpus of Metalinguistic Questions and Answers about English | Accepted to ACL 2023 | null | null | null | cs.CL | http://creativecommons.org/licenses/by-sa/4.0/ | We present ELQA, a corpus of questions and answers in and about the English
language. Collected from two online forums, the >70k questions (from English
learners and others) cover wide-ranging topics including grammar, meaning,
fluency, and etymology. The answers include descriptions of general properties
of English vocabulary and grammar as well as explanations about specific
(correct and incorrect) usage examples. Unlike most NLP datasets, this corpus
is metalinguistic -- it consists of language about language. As such, it can
facilitate investigations of the metalinguistic capabilities of NLU models, as
well as educational applications in the language learning domain. To study
this, we define a free-form question answering task on our dataset and conduct
evaluations on multiple LLMs (Large Language Models) to analyze their capacity
to generate metalinguistic answers.
| [
{
"version": "v1",
"created": "Sun, 1 May 2022 04:29:50 GMT"
},
{
"version": "v2",
"created": "Mon, 3 Jul 2023 17:42:36 GMT"
}
] | 2023-07-04T00:00:00 | [
[
"Behzad",
"Shabnam",
""
],
[
"Sakaguchi",
"Keisuke",
""
],
[
"Schneider",
"Nathan",
""
],
[
"Zeldes",
"Amir",
""
]
] | TITLE: ELQA: A Corpus of Metalinguistic Questions and Answers about English
ABSTRACT: We present ELQA, a corpus of questions and answers in and about the English
language. Collected from two online forums, the >70k questions (from English
learners and others) cover wide-ranging topics including grammar, meaning,
fluency, and etymology. The answers include descriptions of general properties
of English vocabulary and grammar as well as explanations about specific
(correct and incorrect) usage examples. Unlike most NLP datasets, this corpus
is metalinguistic -- it consists of language about language. As such, it can
facilitate investigations of the metalinguistic capabilities of NLU models, as
well as educational applications in the language learning domain. To study
this, we define a free-form question answering task on our dataset and conduct
evaluations on multiple LLMs (Large Language Models) to analyze their capacity
to generate metalinguistic answers.
| new_dataset | 0.712445 |
2206.00667 | Bishwamittra Ghosh | Bishwamittra Ghosh, Debabrota Basu, Kuldeep S. Meel | How Biased are Your Features?: Computing Fairness Influence Functions
with Global Sensitivity Analysis | Proceedings of FAccT, 2023 | null | 10.1145/3593013.3593983 | null | cs.LG cs.AI | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Fairness in machine learning has attained significant focus due to the
widespread application in high-stake decision-making tasks. Unregulated machine
learning classifiers can exhibit bias towards certain demographic groups in
data, thus the quantification and mitigation of classifier bias is a central
concern in fairness in machine learning. In this paper, we aim to quantify the
influence of different features in a dataset on the bias of a classifier. To do
this, we introduce the Fairness Influence Function (FIF). This function breaks
down bias into its components among individual features and the intersection of
multiple features. The key idea is to represent existing group fairness metrics
as the difference of the scaled conditional variances in the classifier's
prediction and apply a decomposition of variance according to global
sensitivity analysis. To estimate FIFs, we instantiate an algorithm
FairXplainer that applies variance decomposition of classifier's prediction
following local regression. Experiments demonstrate that FairXplainer captures
FIFs of individual feature and intersectional features, provides a better
approximation of bias based on FIFs, demonstrates higher correlation of FIFs
with fairness interventions, and detects changes in bias due to fairness
affirmative/punitive actions in the classifier.
The code is available at https://github.com/ReAILe/bias-explainer.
| [
{
"version": "v1",
"created": "Wed, 1 Jun 2022 04:02:16 GMT"
},
{
"version": "v2",
"created": "Fri, 10 Feb 2023 04:03:59 GMT"
},
{
"version": "v3",
"created": "Mon, 3 Jul 2023 00:07:10 GMT"
}
] | 2023-07-04T00:00:00 | [
[
"Ghosh",
"Bishwamittra",
""
],
[
"Basu",
"Debabrota",
""
],
[
"Meel",
"Kuldeep S.",
""
]
] | TITLE: How Biased are Your Features?: Computing Fairness Influence Functions
with Global Sensitivity Analysis
ABSTRACT: Fairness in machine learning has attained significant focus due to the
widespread application in high-stake decision-making tasks. Unregulated machine
learning classifiers can exhibit bias towards certain demographic groups in
data, thus the quantification and mitigation of classifier bias is a central
concern in fairness in machine learning. In this paper, we aim to quantify the
influence of different features in a dataset on the bias of a classifier. To do
this, we introduce the Fairness Influence Function (FIF). This function breaks
down bias into its components among individual features and the intersection of
multiple features. The key idea is to represent existing group fairness metrics
as the difference of the scaled conditional variances in the classifier's
prediction and apply a decomposition of variance according to global
sensitivity analysis. To estimate FIFs, we instantiate an algorithm
FairXplainer that applies variance decomposition of classifier's prediction
following local regression. Experiments demonstrate that FairXplainer captures
FIFs of individual feature and intersectional features, provides a better
approximation of bias based on FIFs, demonstrates higher correlation of FIFs
with fairness interventions, and detects changes in bias due to fairness
affirmative/punitive actions in the classifier.
The code is available at https://github.com/ReAILe/bias-explainer.
| no_new_dataset | 0.709831 |
2206.03198 | Adam Subel | Adam Subel, Yifei Guan, Ashesh Chattopadhyay, Pedram Hassanzadeh | Explaining the physics of transfer learning a data-driven subgrid-scale
closure to a different turbulent flow | 21 pages, 6 figures | null | 10.1093/pnasnexus/pgad015 | null | physics.flu-dyn cs.LG physics.comp-ph | http://creativecommons.org/licenses/by/4.0/ | Transfer learning (TL) is becoming a powerful tool in scientific applications
of neural networks (NNs), such as weather/climate prediction and turbulence
modeling. TL enables out-of-distribution generalization (e.g., extrapolation in
parameters) and effective blending of disparate training sets (e.g.,
simulations and observations). In TL, selected layers of a NN, already trained
for a base system, are re-trained using a small dataset from a target system.
For effective TL, we need to know 1) what are the best layers to re-train? and
2) what physics are learned during TL? Here, we present novel analyses and a
new framework to address (1)-(2) for a broad range of multi-scale, nonlinear
systems. Our approach combines spectral analyses of the systems' data with
spectral analyses of convolutional NN's activations and kernels, explaining the
inner-workings of TL in terms of the system's nonlinear physics. Using
subgrid-scale modeling of several setups of 2D turbulence as test cases, we
show that the learned kernels are combinations of low-, band-, and high-pass
filters, and that TL learns new filters whose nature is consistent with the
spectral differences of base and target systems. We also find the shallowest
layers are the best to re-train in these cases, which is against the common
wisdom guiding TL in machine learning literature. Our framework identifies the
best layer(s) to re-train beforehand, based on physics and NN theory. Together,
these analyses explain the physics learned in TL and provide a framework to
guide TL for wide-ranging applications in science and engineering, such as
climate change modeling.
| [
{
"version": "v1",
"created": "Tue, 7 Jun 2022 11:42:29 GMT"
}
] | 2023-07-04T00:00:00 | [
[
"Subel",
"Adam",
""
],
[
"Guan",
"Yifei",
""
],
[
"Chattopadhyay",
"Ashesh",
""
],
[
"Hassanzadeh",
"Pedram",
""
]
] | TITLE: Explaining the physics of transfer learning a data-driven subgrid-scale
closure to a different turbulent flow
ABSTRACT: Transfer learning (TL) is becoming a powerful tool in scientific applications
of neural networks (NNs), such as weather/climate prediction and turbulence
modeling. TL enables out-of-distribution generalization (e.g., extrapolation in
parameters) and effective blending of disparate training sets (e.g.,
simulations and observations). In TL, selected layers of a NN, already trained
for a base system, are re-trained using a small dataset from a target system.
For effective TL, we need to know 1) what are the best layers to re-train? and
2) what physics are learned during TL? Here, we present novel analyses and a
new framework to address (1)-(2) for a broad range of multi-scale, nonlinear
systems. Our approach combines spectral analyses of the systems' data with
spectral analyses of convolutional NN's activations and kernels, explaining the
inner-workings of TL in terms of the system's nonlinear physics. Using
subgrid-scale modeling of several setups of 2D turbulence as test cases, we
show that the learned kernels are combinations of low-, band-, and high-pass
filters, and that TL learns new filters whose nature is consistent with the
spectral differences of base and target systems. We also find the shallowest
layers are the best to re-train in these cases, which is against the common
wisdom guiding TL in machine learning literature. Our framework identifies the
best layer(s) to re-train beforehand, based on physics and NN theory. Together,
these analyses explain the physics learned in TL and provide a framework to
guide TL for wide-ranging applications in science and engineering, such as
climate change modeling.
| no_new_dataset | 0.711844 |
2206.07786 | Raed Al Kontar | Xubo Yue, Raed Al Kontar, Ana Mar\'ia Estrada G\'omez | Federated Data Analytics: A Study on Linear Models | null | IISE Transactions, 2023 | 10.1080/24725854.2022.2157912 | null | stat.AP cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | As edge devices become increasingly powerful, data analytics are gradually
moving from a centralized to a decentralized regime where edge compute
resources are exploited to process more of the data locally. This regime of
analytics is coined as federated data analytics (FDA). In spite of the recent
success stories of FDA, most literature focuses exclusively on deep neural
networks. In this work, we take a step back to develop an FDA treatment for one
of the most fundamental statistical models: linear regression. Our treatment is
built upon hierarchical modeling that allows borrowing strength across multiple
groups. To this end, we propose two federated hierarchical model structures
that provide a shared representation across devices to facilitate information
sharing. Notably, our proposed frameworks are capable of providing uncertainty
quantification, variable selection, hypothesis testing and fast adaptation to
new unseen data. We validate our methods on a range of real-life applications
including condition monitoring for aircraft engines. The results show that our
FDA treatment for linear models can serve as a competing benchmark model for
future development of federated algorithms.
| [
{
"version": "v1",
"created": "Wed, 15 Jun 2022 19:50:07 GMT"
}
] | 2023-07-04T00:00:00 | [
[
"Yue",
"Xubo",
""
],
[
"Kontar",
"Raed Al",
""
],
[
"Gómez",
"Ana María Estrada",
""
]
] | TITLE: Federated Data Analytics: A Study on Linear Models
ABSTRACT: As edge devices become increasingly powerful, data analytics are gradually
moving from a centralized to a decentralized regime where edge compute
resources are exploited to process more of the data locally. This regime of
analytics is coined as federated data analytics (FDA). In spite of the recent
success stories of FDA, most literature focuses exclusively on deep neural
networks. In this work, we take a step back to develop an FDA treatment for one
of the most fundamental statistical models: linear regression. Our treatment is
built upon hierarchical modeling that allows borrowing strength across multiple
groups. To this end, we propose two federated hierarchical model structures
that provide a shared representation across devices to facilitate information
sharing. Notably, our proposed frameworks are capable of providing uncertainty
quantification, variable selection, hypothesis testing and fast adaptation to
new unseen data. We validate our methods on a range of real-life applications
including condition monitoring for aircraft engines. The results show that our
FDA treatment for linear models can serve as a competing benchmark model for
future development of federated algorithms.
| no_new_dataset | 0.709221 |
2206.14451 | Yining Shi | Yining Shi, Jingyan Shen, Yifan Sun, Yunlong Wang, Jiaxin Li, Shiqi
Sun, Kun Jiang, Diange Yang | SRCN3D: Sparse R-CNN 3D for Compact Convolutional Multi-View 3D Object
Detection and Tracking | Accepted to Vision-centric Autonomous Driving(VCAD) Workshop at
CVPR2023, For more details refer to http://vcad.site/#/papers | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Detection and tracking of moving objects is an essential component in
environmental perception for autonomous driving. In the flourishing field of
multi-view 3D camera-based detectors, different transformer-based pipelines are
designed to learn queries in 3D space from 2D feature maps of perspective
views, but the dominant dense BEV query mechanism is computationally
inefficient. This paper proposes Sparse R-CNN 3D (SRCN3D), a novel two-stage
fully-sparse detector that incorporates sparse queries, sparse attention with
box-wise sampling, and sparse prediction. SRCN3D adopts a cascade structure
with the twin-track update of both a fixed number of query boxes and latent
query features. Our novel sparse feature sampling module only utilizes local 2D
region of interest (RoI) features calculated by the projection of 3D query
boxes for further box refinement, leading to a fully-convolutional and
deployment-friendly pipeline. For multi-object tracking, motion features, query
features and RoI features are comprehensively utilized in multi-hypotheses data
association. Extensive experiments on nuScenes dataset demonstrate that SRCN3D
achieves competitive performance in both 3D object detection and multi-object
tracking tasks, while also exhibiting superior efficiency compared to
transformer-based methods. Code and models are available at
https://github.com/synsin0/SRCN3D.
| [
{
"version": "v1",
"created": "Wed, 29 Jun 2022 07:58:39 GMT"
},
{
"version": "v2",
"created": "Fri, 21 Oct 2022 04:05:32 GMT"
},
{
"version": "v3",
"created": "Sun, 2 Jul 2023 01:11:12 GMT"
}
] | 2023-07-04T00:00:00 | [
[
"Shi",
"Yining",
""
],
[
"Shen",
"Jingyan",
""
],
[
"Sun",
"Yifan",
""
],
[
"Wang",
"Yunlong",
""
],
[
"Li",
"Jiaxin",
""
],
[
"Sun",
"Shiqi",
""
],
[
"Jiang",
"Kun",
""
],
[
"Yang",
"Diange",
""
]
] | TITLE: SRCN3D: Sparse R-CNN 3D for Compact Convolutional Multi-View 3D Object
Detection and Tracking
ABSTRACT: Detection and tracking of moving objects is an essential component in
environmental perception for autonomous driving. In the flourishing field of
multi-view 3D camera-based detectors, different transformer-based pipelines are
designed to learn queries in 3D space from 2D feature maps of perspective
views, but the dominant dense BEV query mechanism is computationally
inefficient. This paper proposes Sparse R-CNN 3D (SRCN3D), a novel two-stage
fully-sparse detector that incorporates sparse queries, sparse attention with
box-wise sampling, and sparse prediction. SRCN3D adopts a cascade structure
with the twin-track update of both a fixed number of query boxes and latent
query features. Our novel sparse feature sampling module only utilizes local 2D
region of interest (RoI) features calculated by the projection of 3D query
boxes for further box refinement, leading to a fully-convolutional and
deployment-friendly pipeline. For multi-object tracking, motion features, query
features and RoI features are comprehensively utilized in multi-hypotheses data
association. Extensive experiments on nuScenes dataset demonstrate that SRCN3D
achieves competitive performance in both 3D object detection and multi-object
tracking tasks, while also exhibiting superior efficiency compared to
transformer-based methods. Code and models are available at
https://github.com/synsin0/SRCN3D.
| no_new_dataset | 0.707967 |
2207.01518 | Fengyuan Liu | Fengyuan Liu, Petter Holme, Matteo Chiesa, Bedoor AlShebli, Talal
Rahwan | Gender inequality and self-publication patterns among scientific editors | null | Nat Hum Behav 7, 353-364 (2023) | 10.1038/s41562-022-01498-1 | null | cs.CY cs.DL | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Academic publishing is the principal medium of documenting and disseminating
scientific discoveries. At the heart of its daily operations are the editorial
boards. Despite their activities and recruitment often being opaque to outside
observers, they play a crucial role in promoting fair evaluations and gender
parity. Literature on gender inequality lacks the connection between women as
editors and as research-active scientists, thereby missing the comparison
between the gender balances in these two academic roles. Literature on
editorial fairness similarly lacks longitudinal studies on the conflicts of
interest arising from editors being research active, which motivates them to
expedite the publication of their papers. We fill these gaps using a dataset of
103,000 editors, 240 million authors, and 220 million publications spanning
five decades and 15 disciplines. This unique dataset allows us to compare the
proportion of female editors to that of female scientists in any given year or
discipline. Although women are already underrepresented in science (26%), they
are even more so among editors (14%) and editors-in-chief (8%); the lack of
women with long-enough publishing careers explains the gender gap among
editors, but not editors-in-chief, suggesting that other factors may be at
play. Our dataset also allows us to study the self-publication patterns of
editors, revealing that 8% of them double the rate at which they publish in
their own journal soon after the editorship starts, and this behavior is
accentuated in journals where the editors-in-chief self-publish excessively.
Finally, men are more likely to engage in this behaviour than women.
| [
{
"version": "v1",
"created": "Thu, 23 Jun 2022 10:22:01 GMT"
}
] | 2023-07-04T00:00:00 | [
[
"Liu",
"Fengyuan",
""
],
[
"Holme",
"Petter",
""
],
[
"Chiesa",
"Matteo",
""
],
[
"AlShebli",
"Bedoor",
""
],
[
"Rahwan",
"Talal",
""
]
] | TITLE: Gender inequality and self-publication patterns among scientific editors
ABSTRACT: Academic publishing is the principal medium of documenting and disseminating
scientific discoveries. At the heart of its daily operations are the editorial
boards. Despite their activities and recruitment often being opaque to outside
observers, they play a crucial role in promoting fair evaluations and gender
parity. Literature on gender inequality lacks the connection between women as
editors and as research-active scientists, thereby missing the comparison
between the gender balances in these two academic roles. Literature on
editorial fairness similarly lacks longitudinal studies on the conflicts of
interest arising from editors being research active, which motivates them to
expedite the publication of their papers. We fill these gaps using a dataset of
103,000 editors, 240 million authors, and 220 million publications spanning
five decades and 15 disciplines. This unique dataset allows us to compare the
proportion of female editors to that of female scientists in any given year or
discipline. Although women are already underrepresented in science (26%), they
are even more so among editors (14%) and editors-in-chief (8%); the lack of
women with long-enough publishing careers explains the gender gap among
editors, but not editors-in-chief, suggesting that other factors may be at
play. Our dataset also allows us to study the self-publication patterns of
editors, revealing that 8% of them double the rate at which they publish in
their own journal soon after the editorship starts, and this behavior is
accentuated in journals where the editors-in-chief self-publish excessively.
Finally, men are more likely to engage in this behaviour than women.
| new_dataset | 0.714223 |
2207.10553 | Jennifer J. Sun | Jennifer J. Sun, Markus Marks, Andrew Ulmer, Dipam Chakraborty, Brian
Geuther, Edward Hayes, Heng Jia, Vivek Kumar, Sebastian Oleszko, Zachary
Partridge, Milan Peelman, Alice Robie, Catherine E. Schretter, Keith
Sheppard, Chao Sun, Param Uttarwar, Julian M. Wagner, Eric Werner, Joseph
Parker, Pietro Perona, Yisong Yue, Kristin Branson, Ann Kennedy | MABe22: A Multi-Species Multi-Task Benchmark for Learned Representations
of Behavior | To appear in ICML 2023, Project website:
https://sites.google.com/view/computational-behavior/our-datasets/mabe2022-dataset | null | null | null | cs.LG cs.AI cs.CV cs.MA | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We introduce MABe22, a large-scale, multi-agent video and trajectory
benchmark to assess the quality of learned behavior representations. This
dataset is collected from a variety of biology experiments, and includes
triplets of interacting mice (4.7 million frames video+pose tracking data, 10
million frames pose only), symbiotic beetle-ant interactions (10 million frames
video data), and groups of interacting flies (4.4 million frames of pose
tracking data). Accompanying these data, we introduce a panel of real-life
downstream analysis tasks to assess the quality of learned representations by
evaluating how well they preserve information about the experimental conditions
(e.g. strain, time of day, optogenetic stimulation) and animal behavior. We
test multiple state-of-the-art self-supervised video and trajectory
representation learning methods to demonstrate the use of our benchmark,
revealing that methods developed using human action datasets do not fully
translate to animal datasets. We hope that our benchmark and dataset encourage
a broader exploration of behavior representation learning methods across
species and settings.
| [
{
"version": "v1",
"created": "Thu, 21 Jul 2022 15:51:30 GMT"
},
{
"version": "v2",
"created": "Fri, 30 Jun 2023 22:45:47 GMT"
}
] | 2023-07-04T00:00:00 | [
[
"Sun",
"Jennifer J.",
""
],
[
"Marks",
"Markus",
""
],
[
"Ulmer",
"Andrew",
""
],
[
"Chakraborty",
"Dipam",
""
],
[
"Geuther",
"Brian",
""
],
[
"Hayes",
"Edward",
""
],
[
"Jia",
"Heng",
""
],
[
"Kumar",
"Vivek",
""
],
[
"Oleszko",
"Sebastian",
""
],
[
"Partridge",
"Zachary",
""
],
[
"Peelman",
"Milan",
""
],
[
"Robie",
"Alice",
""
],
[
"Schretter",
"Catherine E.",
""
],
[
"Sheppard",
"Keith",
""
],
[
"Sun",
"Chao",
""
],
[
"Uttarwar",
"Param",
""
],
[
"Wagner",
"Julian M.",
""
],
[
"Werner",
"Eric",
""
],
[
"Parker",
"Joseph",
""
],
[
"Perona",
"Pietro",
""
],
[
"Yue",
"Yisong",
""
],
[
"Branson",
"Kristin",
""
],
[
"Kennedy",
"Ann",
""
]
] | TITLE: MABe22: A Multi-Species Multi-Task Benchmark for Learned Representations
of Behavior
ABSTRACT: We introduce MABe22, a large-scale, multi-agent video and trajectory
benchmark to assess the quality of learned behavior representations. This
dataset is collected from a variety of biology experiments, and includes
triplets of interacting mice (4.7 million frames video+pose tracking data, 10
million frames pose only), symbiotic beetle-ant interactions (10 million frames
video data), and groups of interacting flies (4.4 million frames of pose
tracking data). Accompanying these data, we introduce a panel of real-life
downstream analysis tasks to assess the quality of learned representations by
evaluating how well they preserve information about the experimental conditions
(e.g. strain, time of day, optogenetic stimulation) and animal behavior. We
test multiple state-of-the-art self-supervised video and trajectory
representation learning methods to demonstrate the use of our benchmark,
revealing that methods developed using human action datasets do not fully
translate to animal datasets. We hope that our benchmark and dataset encourage
a broader exploration of behavior representation learning methods across
species and settings.
| new_dataset | 0.719029 |
2208.00329 | Sayontan Ghosh | Sayontan Ghosh, Mahnaz Koupaee, Isabella Chen, Francis Ferraro,
Nathanael Chambers, Niranjan Balasubramanian | PASTA: A Dataset for Modeling Participant States in Narratives | null | null | null | null | cs.CL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The events in a narrative are understood as a coherent whole via the
underlying states of their participants. Often, these participant states are
not explicitly mentioned, instead left to be inferred by the reader. A model
that understands narratives should likewise infer these implicit states, and
even reason about the impact of changes to these states on the narrative. To
facilitate this goal, we introduce a new crowdsourced English-language,
Participant States dataset, PASTA. This dataset contains inferable participant
states; a counterfactual perturbation to each state; and the changes to the
story that would be necessary if the counterfactual were true. We introduce
three state-based reasoning tasks that test for the ability to infer when a
state is entailed by a story, to revise a story conditioned on a counterfactual
state, and to explain the most likely state change given a revised story.
Experiments show that today's LLMs can reason about states to some degree, but
there is large room for improvement, especially in problems requiring access
and ability to reason with diverse types of knowledge (e.g. physical,
numerical, factual).
| [
{
"version": "v1",
"created": "Sun, 31 Jul 2022 01:21:48 GMT"
},
{
"version": "v2",
"created": "Sat, 1 Jul 2023 22:34:52 GMT"
}
] | 2023-07-04T00:00:00 | [
[
"Ghosh",
"Sayontan",
""
],
[
"Koupaee",
"Mahnaz",
""
],
[
"Chen",
"Isabella",
""
],
[
"Ferraro",
"Francis",
""
],
[
"Chambers",
"Nathanael",
""
],
[
"Balasubramanian",
"Niranjan",
""
]
] | TITLE: PASTA: A Dataset for Modeling Participant States in Narratives
ABSTRACT: The events in a narrative are understood as a coherent whole via the
underlying states of their participants. Often, these participant states are
not explicitly mentioned, instead left to be inferred by the reader. A model
that understands narratives should likewise infer these implicit states, and
even reason about the impact of changes to these states on the narrative. To
facilitate this goal, we introduce a new crowdsourced English-language,
Participant States dataset, PASTA. This dataset contains inferable participant
states; a counterfactual perturbation to each state; and the changes to the
story that would be necessary if the counterfactual were true. We introduce
three state-based reasoning tasks that test for the ability to infer when a
state is entailed by a story, to revise a story conditioned on a counterfactual
state, and to explain the most likely state change given a revised story.
Experiments show that today's LLMs can reason about states to some degree, but
there is large room for improvement, especially in problems requiring access
and ability to reason with diverse types of knowledge (e.g. physical,
numerical, factual).
| new_dataset | 0.717055 |
2208.06146 | Trent Henderson | Trent Henderson and Ben D. Fulcher | Feature-Based Time-Series Analysis in R using the theft Package | null | null | null | null | stat.ML cs.LG cs.MS q-bio.QM stat.AP stat.ME | http://creativecommons.org/licenses/by/4.0/ | Time series are measured and analyzed across the sciences. One method of
quantifying the structure of time series is by calculating a set of summary
statistics or `features', and then representing a time series in terms of its
properties as a feature vector. The resulting feature space is interpretable
and informative, and enables conventional statistical learning approaches,
including clustering, regression, and classification, to be applied to
time-series datasets. Many open-source software packages for computing sets of
time-series features exist across multiple programming languages, including
catch22 (22 features: Matlab, R, Python, Julia), feasts (42 features: R),
tsfeatures (63 features: R), Kats (40 features: Python), tsfresh (779 features:
Python), and TSFEL (390 features: Python). However, there are several issues:
(i) a singular access point to these packages is not currently available; (ii)
to access all feature sets, users must be fluent in multiple languages; and
(iii) these feature-extraction packages lack extensive accompanying
methodological pipelines for performing feature-based time-series analysis,
such as applications to time-series classification. Here we introduce a
solution to these issues in an R software package called theft: Tools for
Handling Extraction of Features from Time series. theft is a unified and
extendable framework for computing features from the six open-source
time-series feature sets listed above. It also includes a suite of functions
for processing and interpreting the performance of extracted features,
including extensive data-visualization templates, low-dimensional projections,
and time-series classification operations. With an increasing volume and
complexity of time-series datasets in the sciences and industry, theft provides
a standardized framework for comprehensively quantifying and interpreting
informative structure in time series.
| [
{
"version": "v1",
"created": "Fri, 12 Aug 2022 07:29:29 GMT"
},
{
"version": "v2",
"created": "Mon, 15 Aug 2022 23:23:36 GMT"
},
{
"version": "v3",
"created": "Wed, 17 Aug 2022 01:07:55 GMT"
},
{
"version": "v4",
"created": "Mon, 3 Jul 2023 09:02:40 GMT"
}
] | 2023-07-04T00:00:00 | [
[
"Henderson",
"Trent",
""
],
[
"Fulcher",
"Ben D.",
""
]
] | TITLE: Feature-Based Time-Series Analysis in R using the theft Package
ABSTRACT: Time series are measured and analyzed across the sciences. One method of
quantifying the structure of time series is by calculating a set of summary
statistics or `features', and then representing a time series in terms of its
properties as a feature vector. The resulting feature space is interpretable
and informative, and enables conventional statistical learning approaches,
including clustering, regression, and classification, to be applied to
time-series datasets. Many open-source software packages for computing sets of
time-series features exist across multiple programming languages, including
catch22 (22 features: Matlab, R, Python, Julia), feasts (42 features: R),
tsfeatures (63 features: R), Kats (40 features: Python), tsfresh (779 features:
Python), and TSFEL (390 features: Python). However, there are several issues:
(i) a singular access point to these packages is not currently available; (ii)
to access all feature sets, users must be fluent in multiple languages; and
(iii) these feature-extraction packages lack extensive accompanying
methodological pipelines for performing feature-based time-series analysis,
such as applications to time-series classification. Here we introduce a
solution to these issues in an R software package called theft: Tools for
Handling Extraction of Features from Time series. theft is a unified and
extendable framework for computing features from the six open-source
time-series feature sets listed above. It also includes a suite of functions
for processing and interpreting the performance of extracted features,
including extensive data-visualization templates, low-dimensional projections,
and time-series classification operations. With an increasing volume and
complexity of time-series datasets in the sciences and industry, theft provides
a standardized framework for comprehensively quantifying and interpreting
informative structure in time series.
| no_new_dataset | 0.709044 |
2208.13754 | Rotem Liss | Michel Boyer, Gilles Brassard, Nicolas Godbout, Rotem Liss, St\'ephane
Virally | Simple and Rigorous Proof Method for the Security of Practical Quantum
Key Distribution in the Single-Qubit Regime Using Mismatched Basis
Measurements | 25 pages | Quantum Reports 2023, 5(1), 52-77 | 10.3390/quantum5010005 | null | quant-ph cs.CR | http://creativecommons.org/licenses/by/4.0/ | Quantum key distribution (QKD) protocols aim at allowing two parties to
generate a secret shared key. While many QKD protocols have been proven
unconditionally secure in theory, practical security analyses of experimental
QKD implementations typically do not take into account all possible loopholes,
and practical devices are still not fully characterized for obtaining tight and
realistic key rates. We present a simple method of computing secure key rates
for any practical implementation of discrete-variable QKD (which can also apply
to measurement-device-independent QKD), initially in the single-qubit lossless
regime, and we rigorously prove its unconditional security against any possible
attack. We hope our method becomes one of the standard tools used for
analysing, benchmarking, and standardizing all practical realizations of QKD.
| [
{
"version": "v1",
"created": "Mon, 29 Aug 2022 17:37:58 GMT"
},
{
"version": "v2",
"created": "Mon, 28 Nov 2022 17:11:23 GMT"
},
{
"version": "v3",
"created": "Mon, 3 Jul 2023 16:55:26 GMT"
}
] | 2023-07-04T00:00:00 | [
[
"Boyer",
"Michel",
""
],
[
"Brassard",
"Gilles",
""
],
[
"Godbout",
"Nicolas",
""
],
[
"Liss",
"Rotem",
""
],
[
"Virally",
"Stéphane",
""
]
] | TITLE: Simple and Rigorous Proof Method for the Security of Practical Quantum
Key Distribution in the Single-Qubit Regime Using Mismatched Basis
Measurements
ABSTRACT: Quantum key distribution (QKD) protocols aim at allowing two parties to
generate a secret shared key. While many QKD protocols have been proven
unconditionally secure in theory, practical security analyses of experimental
QKD implementations typically do not take into account all possible loopholes,
and practical devices are still not fully characterized for obtaining tight and
realistic key rates. We present a simple method of computing secure key rates
for any practical implementation of discrete-variable QKD (which can also apply
to measurement-device-independent QKD), initially in the single-qubit lossless
regime, and we rigorously prove its unconditional security against any possible
attack. We hope our method becomes one of the standard tools used for
analysing, benchmarking, and standardizing all practical realizations of QKD.
| no_new_dataset | 0.708799 |
2209.01599 | Yanna Lin | Yanna Lin, Haotian Li, Aoyu Wu, Yong Wang, Huamin Qu | DMiner: Dashboard Design Mining and Recommendation | in IEEE Transactions on Visualization and Computer Graphics | null | 10.1109/TVCG.2023.3251344 | null | cs.HC cs.GR | http://creativecommons.org/licenses/by/4.0/ | Dashboards, which comprise multiple views on a single display, help analyze
and communicate multiple perspectives of data simultaneously. However, creating
effective and elegant dashboards is challenging since it requires careful and
logical arrangement and coordination of multiple visualizations. To solve the
problem, we propose a data-driven approach for mining design rules from
dashboards and automating dashboard organization. Specifically, we focus on two
prominent aspects of the organization: arrangement, which describes the
position, size, and layout of each view in the display space; and coordination,
which indicates the interaction between pairwise views. We build a new dataset
containing 854 dashboards crawled online, and develop feature engineering
methods for describing the single views and view-wise relationships in terms of
data, encoding, layout, and interactions. Further, we identify design rules
among those features and develop a recommender for dashboard design. We
demonstrate the usefulness of DMiner through an expert study and a user study.
The expert study shows that our extracted design rules are reasonable and
conform to the design practice of experts. Moreover, a comparative user study
shows that our recommender could help automate dashboard organization and reach
human-level performance. In summary, our work offers a promising starting point
for design mining visualizations to build recommenders.
| [
{
"version": "v1",
"created": "Sun, 4 Sep 2022 11:46:02 GMT"
},
{
"version": "v2",
"created": "Fri, 24 Feb 2023 07:32:56 GMT"
}
] | 2023-07-04T00:00:00 | [
[
"Lin",
"Yanna",
""
],
[
"Li",
"Haotian",
""
],
[
"Wu",
"Aoyu",
""
],
[
"Wang",
"Yong",
""
],
[
"Qu",
"Huamin",
""
]
] | TITLE: DMiner: Dashboard Design Mining and Recommendation
ABSTRACT: Dashboards, which comprise multiple views on a single display, help analyze
and communicate multiple perspectives of data simultaneously. However, creating
effective and elegant dashboards is challenging since it requires careful and
logical arrangement and coordination of multiple visualizations. To solve the
problem, we propose a data-driven approach for mining design rules from
dashboards and automating dashboard organization. Specifically, we focus on two
prominent aspects of the organization: arrangement, which describes the
position, size, and layout of each view in the display space; and coordination,
which indicates the interaction between pairwise views. We build a new dataset
containing 854 dashboards crawled online, and develop feature engineering
methods for describing the single views and view-wise relationships in terms of
data, encoding, layout, and interactions. Further, we identify design rules
among those features and develop a recommender for dashboard design. We
demonstrate the usefulness of DMiner through an expert study and a user study.
The expert study shows that our extracted design rules are reasonable and
conform to the design practice of experts. Moreover, a comparative user study
shows that our recommender could help automate dashboard organization and reach
human-level performance. In summary, our work offers a promising starting point
for design mining visualizations to build recommenders.
| new_dataset | 0.713007 |
2209.03475 | Suman Kunwar | Suman Kunwar | Convolutional Neural Network (CNN) to reduce construction loss in JPEG
compression caused by Discrete Fourier Transform (DFT) | null | null | null | null | eess.IV cs.CV cs.LG | http://creativecommons.org/licenses/by/4.0/ | In recent decades, digital image processing has gained enormous popularity.
Consequently, a number of data compression strategies have been put forth, with
the goal of minimizing the amount of information required to represent images.
Among them, JPEG compression is one of the most popular methods that has been
widely applied in multimedia and digital applications. The periodic nature of
DFT makes it impossible to meet the periodic condition of an image's opposing
edges without producing severe artifacts, which lowers the image's perceptual
visual quality. On the other hand, deep learning has recently achieved
outstanding results for applications like speech recognition, image reduction,
and natural language processing. Convolutional Neural Networks (CNN) have
received more attention than most other types of deep neural networks. The use
of convolution in feature extraction results in a less redundant feature map
and a smaller dataset, both of which are crucial for image compression. In this
work, an effective image compression method is purposed using autoencoders. The
study's findings revealed a number of important trends that suggested better
reconstruction along with good compression can be achieved using autoencoders.
| [
{
"version": "v1",
"created": "Fri, 26 Aug 2022 12:46:16 GMT"
},
{
"version": "v2",
"created": "Sun, 2 Jul 2023 23:41:55 GMT"
}
] | 2023-07-04T00:00:00 | [
[
"Kunwar",
"Suman",
""
]
] | TITLE: Convolutional Neural Network (CNN) to reduce construction loss in JPEG
compression caused by Discrete Fourier Transform (DFT)
ABSTRACT: In recent decades, digital image processing has gained enormous popularity.
Consequently, a number of data compression strategies have been put forth, with
the goal of minimizing the amount of information required to represent images.
Among them, JPEG compression is one of the most popular methods that has been
widely applied in multimedia and digital applications. The periodic nature of
DFT makes it impossible to meet the periodic condition of an image's opposing
edges without producing severe artifacts, which lowers the image's perceptual
visual quality. On the other hand, deep learning has recently achieved
outstanding results for applications like speech recognition, image reduction,
and natural language processing. Convolutional Neural Networks (CNN) have
received more attention than most other types of deep neural networks. The use
of convolution in feature extraction results in a less redundant feature map
and a smaller dataset, both of which are crucial for image compression. In this
work, an effective image compression method is purposed using autoencoders. The
study's findings revealed a number of important trends that suggested better
reconstruction along with good compression can be achieved using autoencoders.
| no_new_dataset | 0.708572 |
2209.13678 | Teresa Salazar | Teresa Salazar, Miguel Fernandes, Helder Araujo, Pedro Henriques Abreu | FAIR-FATE: Fair Federated Learning with Momentum | This preprint has not undergone peer review or any post-submission
improvements or corrections. The Version of Record of this contribution is
published in ICCS 2023 - Lecture Notes in Computer Science, vol 14073,
Springer, and is available online at
https://doi.org/10.1007/978-3-031-35995-8_37 | Computational Science - ICCS 2023. ICCS 2023. Lecture Notes in
Computer Science, vol 14073. Springer, Cham | 10.1007/978-3-031-35995-8_37 | null | cs.LG cs.AI cs.DC | http://creativecommons.org/licenses/by/4.0/ | While fairness-aware machine learning algorithms have been receiving
increasing attention, the focus has been on centralized machine learning,
leaving decentralized methods underexplored. Federated Learning is a
decentralized form of machine learning where clients train local models with a
server aggregating them to obtain a shared global model. Data heterogeneity
amongst clients is a common characteristic of Federated Learning, which may
induce or exacerbate discrimination of unprivileged groups defined by sensitive
attributes such as race or gender. In this work we propose FAIR-FATE: a novel
FAIR FederATEd Learning algorithm that aims to achieve group fairness while
maintaining high utility via a fairness-aware aggregation method that computes
the global model by taking into account the fairness of the clients. To achieve
that, the global model update is computed by estimating a fair model update
using a Momentum term that helps to overcome the oscillations of non-fair
gradients. To the best of our knowledge, this is the first approach in machine
learning that aims to achieve fairness using a fair Momentum estimate.
Experimental results on real-world datasets demonstrate that FAIR-FATE
outperforms state-of-the-art fair Federated Learning algorithms under different
levels of data heterogeneity.
| [
{
"version": "v1",
"created": "Tue, 27 Sep 2022 20:33:38 GMT"
},
{
"version": "v2",
"created": "Sun, 2 Jul 2023 18:35:14 GMT"
}
] | 2023-07-04T00:00:00 | [
[
"Salazar",
"Teresa",
""
],
[
"Fernandes",
"Miguel",
""
],
[
"Araujo",
"Helder",
""
],
[
"Abreu",
"Pedro Henriques",
""
]
] | TITLE: FAIR-FATE: Fair Federated Learning with Momentum
ABSTRACT: While fairness-aware machine learning algorithms have been receiving
increasing attention, the focus has been on centralized machine learning,
leaving decentralized methods underexplored. Federated Learning is a
decentralized form of machine learning where clients train local models with a
server aggregating them to obtain a shared global model. Data heterogeneity
amongst clients is a common characteristic of Federated Learning, which may
induce or exacerbate discrimination of unprivileged groups defined by sensitive
attributes such as race or gender. In this work we propose FAIR-FATE: a novel
FAIR FederATEd Learning algorithm that aims to achieve group fairness while
maintaining high utility via a fairness-aware aggregation method that computes
the global model by taking into account the fairness of the clients. To achieve
that, the global model update is computed by estimating a fair model update
using a Momentum term that helps to overcome the oscillations of non-fair
gradients. To the best of our knowledge, this is the first approach in machine
learning that aims to achieve fairness using a fair Momentum estimate.
Experimental results on real-world datasets demonstrate that FAIR-FATE
outperforms state-of-the-art fair Federated Learning algorithms under different
levels of data heterogeneity.
| no_new_dataset | 0.710208 |
2209.15635 | Wenjie Li | Wenjie Li, Qiaolin Xia, Hao Cheng, Kouyin Xue, Shu-Tao Xia | Vertical Semi-Federated Learning for Efficient Online Advertising | This version is accepted to the Federated Learning workshop of
IJCAI2023 (FL-IJCAI23) | null | null | null | cs.LG cs.IR | http://creativecommons.org/licenses/by/4.0/ | The traditional vertical federated learning schema suffers from two main
issues: 1) restricted applicable scope to overlapped samples and 2) high system
challenge of real-time federated serving, which limits its application to
advertising systems. To this end, we advocate a new learning setting Semi-VFL
(Vertical Semi-Federated Learning) to tackle these challenge. Semi-VFL is
proposed to achieve a practical industry application fashion for VFL, by
learning a federation-aware local model which performs better than single-party
models and meanwhile maintain the convenience of local-serving. For this
purpose, we propose the carefully designed Joint Privileged Learning framework
(JPL) to i) alleviate the absence of the passive party's feature and ii) adapt
to the whole sample space. Specifically, we build an inference-efficient
single-party student model applicable to the whole sample space and meanwhile
maintain the advantage of the federated feature extension. New representation
distillation methods are designed to extract cross-party feature correlations
for both the overlapped and non-overlapped data. We conducted extensive
experiments on real-world advertising datasets. The results show that our
method achieves the best performance over baseline methods and validate its
superiority in the Semi-VFL setting.
| [
{
"version": "v1",
"created": "Fri, 30 Sep 2022 17:59:27 GMT"
},
{
"version": "v2",
"created": "Sat, 1 Jul 2023 14:41:25 GMT"
}
] | 2023-07-04T00:00:00 | [
[
"Li",
"Wenjie",
""
],
[
"Xia",
"Qiaolin",
""
],
[
"Cheng",
"Hao",
""
],
[
"Xue",
"Kouyin",
""
],
[
"Xia",
"Shu-Tao",
""
]
] | TITLE: Vertical Semi-Federated Learning for Efficient Online Advertising
ABSTRACT: The traditional vertical federated learning schema suffers from two main
issues: 1) restricted applicable scope to overlapped samples and 2) high system
challenge of real-time federated serving, which limits its application to
advertising systems. To this end, we advocate a new learning setting Semi-VFL
(Vertical Semi-Federated Learning) to tackle these challenge. Semi-VFL is
proposed to achieve a practical industry application fashion for VFL, by
learning a federation-aware local model which performs better than single-party
models and meanwhile maintain the convenience of local-serving. For this
purpose, we propose the carefully designed Joint Privileged Learning framework
(JPL) to i) alleviate the absence of the passive party's feature and ii) adapt
to the whole sample space. Specifically, we build an inference-efficient
single-party student model applicable to the whole sample space and meanwhile
maintain the advantage of the federated feature extension. New representation
distillation methods are designed to extract cross-party feature correlations
for both the overlapped and non-overlapped data. We conducted extensive
experiments on real-world advertising datasets. The results show that our
method achieves the best performance over baseline methods and validate its
superiority in the Semi-VFL setting.
| no_new_dataset | 0.707859 |
2210.02410 | Adji Bousso Dieng | Dan Friedman and Adji Bousso Dieng | The Vendi Score: A Diversity Evaluation Metric for Machine Learning | The Vendi Score is available as a pip package at
https://github.com/vertaix/Vendi-Score | null | null | null | cs.LG cond-mat.mtrl-sci stat.ML | http://creativecommons.org/licenses/by/4.0/ | Diversity is an important criterion for many areas of machine learning (ML),
including generative modeling and dataset curation. However, existing metrics
for measuring diversity are often domain-specific and limited in flexibility.
In this paper, we address the diversity evaluation problem by proposing the
Vendi Score, which connects and extends ideas from ecology and quantum
statistical mechanics to ML. The Vendi Score is defined as the exponential of
the Shannon entropy of the eigenvalues of a similarity matrix. This matrix is
induced by a user-defined similarity function applied to the sample to be
evaluated for diversity. In taking a similarity function as input, the Vendi
Score enables its user to specify any desired form of diversity. Importantly,
unlike many existing metrics in ML, the Vendi Score does not require a
reference dataset or distribution over samples or labels, it is therefore
general and applicable to any generative model, decoding algorithm, and dataset
from any domain where similarity can be defined. We showcase the Vendi Score on
molecular generative modeling where we found it addresses shortcomings of the
current diversity metric of choice in that domain. We also applied the Vendi
Score to generative models of images and decoding algorithms of text where we
found it confirms known results about diversity in those domains. Furthermore,
we used the Vendi Score to measure mode collapse, a known shortcoming of
generative adversarial networks (GANs). In particular, the Vendi Score revealed
that even GANs that capture all the modes of a labeled dataset can be less
diverse than the original dataset. Finally, the interpretability of the Vendi
Score allowed us to diagnose several benchmark ML datasets for diversity,
opening the door for diversity-informed data augmentation.
| [
{
"version": "v1",
"created": "Wed, 5 Oct 2022 17:32:16 GMT"
},
{
"version": "v2",
"created": "Sun, 2 Jul 2023 22:58:51 GMT"
}
] | 2023-07-04T00:00:00 | [
[
"Friedman",
"Dan",
""
],
[
"Dieng",
"Adji Bousso",
""
]
] | TITLE: The Vendi Score: A Diversity Evaluation Metric for Machine Learning
ABSTRACT: Diversity is an important criterion for many areas of machine learning (ML),
including generative modeling and dataset curation. However, existing metrics
for measuring diversity are often domain-specific and limited in flexibility.
In this paper, we address the diversity evaluation problem by proposing the
Vendi Score, which connects and extends ideas from ecology and quantum
statistical mechanics to ML. The Vendi Score is defined as the exponential of
the Shannon entropy of the eigenvalues of a similarity matrix. This matrix is
induced by a user-defined similarity function applied to the sample to be
evaluated for diversity. In taking a similarity function as input, the Vendi
Score enables its user to specify any desired form of diversity. Importantly,
unlike many existing metrics in ML, the Vendi Score does not require a
reference dataset or distribution over samples or labels, it is therefore
general and applicable to any generative model, decoding algorithm, and dataset
from any domain where similarity can be defined. We showcase the Vendi Score on
molecular generative modeling where we found it addresses shortcomings of the
current diversity metric of choice in that domain. We also applied the Vendi
Score to generative models of images and decoding algorithms of text where we
found it confirms known results about diversity in those domains. Furthermore,
we used the Vendi Score to measure mode collapse, a known shortcoming of
generative adversarial networks (GANs). In particular, the Vendi Score revealed
that even GANs that capture all the modes of a labeled dataset can be less
diverse than the original dataset. Finally, the interpretability of the Vendi
Score allowed us to diagnose several benchmark ML datasets for diversity,
opening the door for diversity-informed data augmentation.
| no_new_dataset | 0.708175 |
2210.04165 | Wei Liu | Wei Liu, Zhilu Lai, Kiran Bacsa, Eleni Chatzi | Neural Extended Kalman Filters for Learning and Predicting Dynamics of
Structural Systems | Accepted for publication in Structural Health Monitoring | Structural Health Monitoring, 2023 | 10.1177/14759217231179912 | null | cs.LG cs.CE nlin.CD stat.ML | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Accurate structural response prediction forms a main driver for structural
health monitoring and control applications. This often requires the proposed
model to adequately capture the underlying dynamics of complex structural
systems. In this work, we utilize a learnable Extended Kalman Filter (EKF),
named the Neural Extended Kalman Filter (Neural EKF) throughout this paper, for
learning the latent evolution dynamics of complex physical systems. The Neural
EKF is a generalized version of the conventional EKF, where the modeling of
process dynamics and sensory observations can be parameterized by neural
networks, therefore learned by end-to-end training. The method is implemented
under the variational inference framework with the EKF conducting inference
from sensing measurements. Typically, conventional variational inference models
are parameterized by neural networks independent of the latent dynamics models.
This characteristic makes the inference and reconstruction accuracy weakly
based on the dynamics models and renders the associated training inadequate. In
this work, we show that the structure imposed by the Neural EKF is beneficial
to the learning process. We demonstrate the efficacy of the framework on both
simulated and real-world structural monitoring datasets, with the results
indicating significant predictive capabilities of the proposed scheme.
| [
{
"version": "v1",
"created": "Sun, 9 Oct 2022 04:39:15 GMT"
},
{
"version": "v2",
"created": "Mon, 3 Jul 2023 09:49:18 GMT"
}
] | 2023-07-04T00:00:00 | [
[
"Liu",
"Wei",
""
],
[
"Lai",
"Zhilu",
""
],
[
"Bacsa",
"Kiran",
""
],
[
"Chatzi",
"Eleni",
""
]
] | TITLE: Neural Extended Kalman Filters for Learning and Predicting Dynamics of
Structural Systems
ABSTRACT: Accurate structural response prediction forms a main driver for structural
health monitoring and control applications. This often requires the proposed
model to adequately capture the underlying dynamics of complex structural
systems. In this work, we utilize a learnable Extended Kalman Filter (EKF),
named the Neural Extended Kalman Filter (Neural EKF) throughout this paper, for
learning the latent evolution dynamics of complex physical systems. The Neural
EKF is a generalized version of the conventional EKF, where the modeling of
process dynamics and sensory observations can be parameterized by neural
networks, therefore learned by end-to-end training. The method is implemented
under the variational inference framework with the EKF conducting inference
from sensing measurements. Typically, conventional variational inference models
are parameterized by neural networks independent of the latent dynamics models.
This characteristic makes the inference and reconstruction accuracy weakly
based on the dynamics models and renders the associated training inadequate. In
this work, we show that the structure imposed by the Neural EKF is beneficial
to the learning process. We demonstrate the efficacy of the framework on both
simulated and real-world structural monitoring datasets, with the results
indicating significant predictive capabilities of the proposed scheme.
| no_new_dataset | 0.71027 |
2210.08710 | Zexian Yang | Zexian Yang, Dayan Wu, Wanqian Zhang, Bo Li, Weiping Wang | Handling Label Uncertainty for Camera Incremental Person
Re-Identification | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Incremental learning for person re-identification (ReID) aims to develop
models that can be trained with a continuous data stream, which is a more
practical setting for real-world applications. However, the existing
incremental ReID methods make two strong assumptions that the cameras are fixed
and the new-emerging data is class-disjoint from previous classes. This is
unrealistic as previously observed pedestrians may re-appear and be captured
again by new cameras. In this paper, we investigate person ReID in an
unexplored scenario named Camera Incremental Person ReID (CIPR), which advances
existing lifelong person ReID by taking into account the class overlap issue.
Specifically, new data collected from new cameras may probably contain an
unknown proportion of identities seen before. This subsequently leads to the
lack of cross-camera annotations for new data due to privacy concerns. To
address these challenges, we propose a novel framework ExtendOVA. First, to
handle the class overlap issue, we introduce an instance-wise seen-class
identification module to discover previously seen identities at the instance
level. Then, we propose a criterion for selecting confident ID-wise candidates
and also devise an early learning regularization term to correct noise issues
in pseudo labels. Furthermore, to compensate for the lack of previous data, we
resort prototypical memory bank to create surrogate features, along with a
cross-camera distillation loss to further retain the inter-camera relationship.
The comprehensive experimental results on multiple benchmarks show that
ExtendOVA significantly outperforms the state-of-the-arts with remarkable
advantages.
| [
{
"version": "v1",
"created": "Mon, 17 Oct 2022 02:59:54 GMT"
},
{
"version": "v2",
"created": "Tue, 18 Oct 2022 02:38:05 GMT"
},
{
"version": "v3",
"created": "Fri, 30 Jun 2023 19:02:52 GMT"
}
] | 2023-07-04T00:00:00 | [
[
"Yang",
"Zexian",
""
],
[
"Wu",
"Dayan",
""
],
[
"Zhang",
"Wanqian",
""
],
[
"Li",
"Bo",
""
],
[
"Wang",
"Weiping",
""
]
] | TITLE: Handling Label Uncertainty for Camera Incremental Person
Re-Identification
ABSTRACT: Incremental learning for person re-identification (ReID) aims to develop
models that can be trained with a continuous data stream, which is a more
practical setting for real-world applications. However, the existing
incremental ReID methods make two strong assumptions that the cameras are fixed
and the new-emerging data is class-disjoint from previous classes. This is
unrealistic as previously observed pedestrians may re-appear and be captured
again by new cameras. In this paper, we investigate person ReID in an
unexplored scenario named Camera Incremental Person ReID (CIPR), which advances
existing lifelong person ReID by taking into account the class overlap issue.
Specifically, new data collected from new cameras may probably contain an
unknown proportion of identities seen before. This subsequently leads to the
lack of cross-camera annotations for new data due to privacy concerns. To
address these challenges, we propose a novel framework ExtendOVA. First, to
handle the class overlap issue, we introduce an instance-wise seen-class
identification module to discover previously seen identities at the instance
level. Then, we propose a criterion for selecting confident ID-wise candidates
and also devise an early learning regularization term to correct noise issues
in pseudo labels. Furthermore, to compensate for the lack of previous data, we
resort prototypical memory bank to create surrogate features, along with a
cross-camera distillation loss to further retain the inter-camera relationship.
The comprehensive experimental results on multiple benchmarks show that
ExtendOVA significantly outperforms the state-of-the-arts with remarkable
advantages.
| no_new_dataset | 0.711055 |
2210.11182 | Hamza Bouzid | Hamza Bouzid, Lahoucine Ballihi | Facial Expression Video Generation Based-On Spatio-temporal
Convolutional GAN: FEV-GAN | 13 pages, 8 figures, accepted in ISWA journal | Intelligent Systems with Applications. 19 (2022) | 10.1016/j.iswa.2022.200139 | null | cs.CV | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Facial expression generation has always been an intriguing task for
scientists and researchers all over the globe. In this context, we present our
novel approach for generating videos of the six basic facial expressions.
Starting from a single neutral facial image and a label indicating the desired
facial expression, we aim to synthesize a video of the given identity
performing the specified facial expression. Our approach, referred to as
FEV-GAN (Facial Expression Video GAN), is based on Spatio-temporal
Convolutional GANs, that are known to model both content and motion in the same
network. Previous methods based on such a network have shown a good ability to
generate coherent videos with smooth temporal evolution. However, they still
suffer from low image quality and low identity preservation capability. In this
work, we address this problem by using a generator composed of two image
encoders. The first one is pre-trained for facial identity feature extraction
and the second for spatial feature extraction. We have qualitatively and
quantitatively evaluated our model on two international facial expression
benchmark databases: MUG and Oulu-CASIA NIR&VIS. The experimental results
analysis demonstrates the effectiveness of our approach in generating videos of
the six basic facial expressions while preserving the input identity. The
analysis also proves that the use of both identity and spatial features
enhances the decoder ability to better preserve the identity and generate
high-quality videos. The code and the pre-trained model will soon be made
publicly available.
| [
{
"version": "v1",
"created": "Thu, 20 Oct 2022 11:54:32 GMT"
}
] | 2023-07-04T00:00:00 | [
[
"Bouzid",
"Hamza",
""
],
[
"Ballihi",
"Lahoucine",
""
]
] | TITLE: Facial Expression Video Generation Based-On Spatio-temporal
Convolutional GAN: FEV-GAN
ABSTRACT: Facial expression generation has always been an intriguing task for
scientists and researchers all over the globe. In this context, we present our
novel approach for generating videos of the six basic facial expressions.
Starting from a single neutral facial image and a label indicating the desired
facial expression, we aim to synthesize a video of the given identity
performing the specified facial expression. Our approach, referred to as
FEV-GAN (Facial Expression Video GAN), is based on Spatio-temporal
Convolutional GANs, that are known to model both content and motion in the same
network. Previous methods based on such a network have shown a good ability to
generate coherent videos with smooth temporal evolution. However, they still
suffer from low image quality and low identity preservation capability. In this
work, we address this problem by using a generator composed of two image
encoders. The first one is pre-trained for facial identity feature extraction
and the second for spatial feature extraction. We have qualitatively and
quantitatively evaluated our model on two international facial expression
benchmark databases: MUG and Oulu-CASIA NIR&VIS. The experimental results
analysis demonstrates the effectiveness of our approach in generating videos of
the six basic facial expressions while preserving the input identity. The
analysis also proves that the use of both identity and spatial features
enhances the decoder ability to better preserve the identity and generate
high-quality videos. The code and the pre-trained model will soon be made
publicly available.
| no_new_dataset | 0.711638 |
2210.16424 | Chao Pan | Chao Pan, Jin Sima, Saurav Prakash, Vishal Rana, Olgica Milenkovic | Machine Unlearning of Federated Clusters | 27 pages. ICLR 2023 | null | null | null | cs.LG cs.CR | http://creativecommons.org/licenses/by-sa/4.0/ | Federated clustering (FC) is an unsupervised learning problem that arises in
a number of practical applications, including personalized recommender and
healthcare systems. With the adoption of recent laws ensuring the "right to be
forgotten", the problem of machine unlearning for FC methods has become of
significant importance. We introduce, for the first time, the problem of
machine unlearning for FC, and propose an efficient unlearning mechanism for a
customized secure FC framework. Our FC framework utilizes special
initialization procedures that we show are well-suited for unlearning. To
protect client data privacy, we develop the secure compressed multiset
aggregation (SCMA) framework that addresses sparse secure federated learning
(FL) problems encountered during clustering as well as more general problems.
To simultaneously facilitate low communication complexity and secret sharing
protocols, we integrate Reed-Solomon encoding with special evaluation points
into our SCMA pipeline, and prove that the client communication cost is
logarithmic in the vector dimension. Additionally, to demonstrate the benefits
of our unlearning mechanism over complete retraining, we provide a theoretical
analysis for the unlearning performance of our approach. Simulation results
show that the new FC framework exhibits superior clustering performance
compared to previously reported FC baselines when the cluster sizes are highly
imbalanced. Compared to completely retraining K-means++ locally and globally
for each removal request, our unlearning procedure offers an average speed-up
of roughly 84x across seven datasets. Our implementation for the proposed
method is available at https://github.com/thupchnsky/mufc.
| [
{
"version": "v1",
"created": "Fri, 28 Oct 2022 22:21:29 GMT"
},
{
"version": "v2",
"created": "Sat, 1 Jul 2023 03:53:30 GMT"
}
] | 2023-07-04T00:00:00 | [
[
"Pan",
"Chao",
""
],
[
"Sima",
"Jin",
""
],
[
"Prakash",
"Saurav",
""
],
[
"Rana",
"Vishal",
""
],
[
"Milenkovic",
"Olgica",
""
]
] | TITLE: Machine Unlearning of Federated Clusters
ABSTRACT: Federated clustering (FC) is an unsupervised learning problem that arises in
a number of practical applications, including personalized recommender and
healthcare systems. With the adoption of recent laws ensuring the "right to be
forgotten", the problem of machine unlearning for FC methods has become of
significant importance. We introduce, for the first time, the problem of
machine unlearning for FC, and propose an efficient unlearning mechanism for a
customized secure FC framework. Our FC framework utilizes special
initialization procedures that we show are well-suited for unlearning. To
protect client data privacy, we develop the secure compressed multiset
aggregation (SCMA) framework that addresses sparse secure federated learning
(FL) problems encountered during clustering as well as more general problems.
To simultaneously facilitate low communication complexity and secret sharing
protocols, we integrate Reed-Solomon encoding with special evaluation points
into our SCMA pipeline, and prove that the client communication cost is
logarithmic in the vector dimension. Additionally, to demonstrate the benefits
of our unlearning mechanism over complete retraining, we provide a theoretical
analysis for the unlearning performance of our approach. Simulation results
show that the new FC framework exhibits superior clustering performance
compared to previously reported FC baselines when the cluster sizes are highly
imbalanced. Compared to completely retraining K-means++ locally and globally
for each removal request, our unlearning procedure offers an average speed-up
of roughly 84x across seven datasets. Our implementation for the proposed
method is available at https://github.com/thupchnsky/mufc.
| no_new_dataset | 0.710804 |
2210.17283 | Mathieu Chevalley | Mathieu Chevalley, Yusuf Roohani, Arash Mehrjou, Jure Leskovec,
Patrick Schwab | CausalBench: A Large-scale Benchmark for Network Inference from
Single-cell Perturbation Data | null | null | null | null | cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Causal inference is a vital aspect of multiple scientific disciplines and is
routinely applied to high-impact applications such as medicine. However,
evaluating the performance of causal inference methods in real-world
environments is challenging due to the need for observations under both
interventional and control conditions. Traditional evaluations conducted on
synthetic datasets do not reflect the performance in real-world systems. To
address this, we introduce CausalBench, a benchmark suite for evaluating
network inference methods on real-world interventional data from large-scale
single-cell perturbation experiments. CausalBench incorporates
biologically-motivated performance metrics, including new distribution-based
interventional metrics. A systematic evaluation of state-of-the-art causal
inference methods using our CausalBench suite highlights how poor scalability
of current methods limits performance. Moreover, methods that use
interventional information do not outperform those that only use observational
data, contrary to what is observed on synthetic benchmarks. Thus, CausalBench
opens new avenues in causal network inference research and provides a
principled and reliable way to track progress in leveraging real-world
interventional data.
| [
{
"version": "v1",
"created": "Mon, 31 Oct 2022 13:04:07 GMT"
},
{
"version": "v2",
"created": "Mon, 3 Jul 2023 09:12:49 GMT"
}
] | 2023-07-04T00:00:00 | [
[
"Chevalley",
"Mathieu",
""
],
[
"Roohani",
"Yusuf",
""
],
[
"Mehrjou",
"Arash",
""
],
[
"Leskovec",
"Jure",
""
],
[
"Schwab",
"Patrick",
""
]
] | TITLE: CausalBench: A Large-scale Benchmark for Network Inference from
Single-cell Perturbation Data
ABSTRACT: Causal inference is a vital aspect of multiple scientific disciplines and is
routinely applied to high-impact applications such as medicine. However,
evaluating the performance of causal inference methods in real-world
environments is challenging due to the need for observations under both
interventional and control conditions. Traditional evaluations conducted on
synthetic datasets do not reflect the performance in real-world systems. To
address this, we introduce CausalBench, a benchmark suite for evaluating
network inference methods on real-world interventional data from large-scale
single-cell perturbation experiments. CausalBench incorporates
biologically-motivated performance metrics, including new distribution-based
interventional metrics. A systematic evaluation of state-of-the-art causal
inference methods using our CausalBench suite highlights how poor scalability
of current methods limits performance. Moreover, methods that use
interventional information do not outperform those that only use observational
data, contrary to what is observed on synthetic benchmarks. Thus, CausalBench
opens new avenues in causal network inference research and provides a
principled and reliable way to track progress in leveraging real-world
interventional data.
| new_dataset | 0.604807 |
2211.02133 | Pingchuan Ma | Pingchuan Ma, Niko Moritz, Stavros Petridis, Christian Fuegen, Maja
Pantic | Streaming Audio-Visual Speech Recognition with Alignment Regularization | Accepted to Interspeech 2023 | null | null | null | eess.AS cs.CV cs.SD | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this work, we propose a streaming AV-ASR system based on a hybrid
connectionist temporal classification (CTC)/attention neural network
architecture. The audio and the visual encoder neural networks are both based
on the conformer architecture, which is made streamable using chunk-wise
self-attention (CSA) and causal convolution. Streaming recognition with a
decoder neural network is realized by using the triggered attention technique,
which performs time-synchronous decoding with joint CTC/attention scoring.
Additionally, we propose a novel alignment regularization technique that
promotes synchronization of the audio and visual encoder, which in turn results
in better word error rates (WERs) at all SNR levels for streaming and offline
AV-ASR models. The proposed AV-ASR model achieves WERs of 2.0% and 2.6% on the
Lip Reading Sentences 3 (LRS3) dataset in an offline and online setup,
respectively, which both present state-of-the-art results when no external
training data are used.
| [
{
"version": "v1",
"created": "Thu, 3 Nov 2022 20:20:47 GMT"
},
{
"version": "v2",
"created": "Sun, 2 Jul 2023 00:33:36 GMT"
}
] | 2023-07-04T00:00:00 | [
[
"Ma",
"Pingchuan",
""
],
[
"Moritz",
"Niko",
""
],
[
"Petridis",
"Stavros",
""
],
[
"Fuegen",
"Christian",
""
],
[
"Pantic",
"Maja",
""
]
] | TITLE: Streaming Audio-Visual Speech Recognition with Alignment Regularization
ABSTRACT: In this work, we propose a streaming AV-ASR system based on a hybrid
connectionist temporal classification (CTC)/attention neural network
architecture. The audio and the visual encoder neural networks are both based
on the conformer architecture, which is made streamable using chunk-wise
self-attention (CSA) and causal convolution. Streaming recognition with a
decoder neural network is realized by using the triggered attention technique,
which performs time-synchronous decoding with joint CTC/attention scoring.
Additionally, we propose a novel alignment regularization technique that
promotes synchronization of the audio and visual encoder, which in turn results
in better word error rates (WERs) at all SNR levels for streaming and offline
AV-ASR models. The proposed AV-ASR model achieves WERs of 2.0% and 2.6% on the
Lip Reading Sentences 3 (LRS3) dataset in an offline and online setup,
respectively, which both present state-of-the-art results when no external
training data are used.
| no_new_dataset | 0.711067 |
2211.03216 | Chao Pan | Chao Pan, Eli Chien, Olgica Milenkovic | Unlearning Graph Classifiers with Limited Data Resources | 12 pages. WWW 2023 | null | null | null | cs.LG cs.CR | http://creativecommons.org/licenses/by-sa/4.0/ | As the demand for user privacy grows, controlled data removal (machine
unlearning) is becoming an important feature of machine learning models for
data-sensitive Web applications such as social networks and recommender
systems. Nevertheless, at this point it is still largely unknown how to perform
efficient machine unlearning of graph neural networks (GNNs); this is
especially the case when the number of training samples is small, in which case
unlearning can seriously compromise the performance of the model. To address
this issue, we initiate the study of unlearning the Graph Scattering Transform
(GST), a mathematical framework that is efficient, provably stable under
feature or graph topology perturbations, and offers graph classification
performance comparable to that of GNNs. Our main contribution is the first
known nonlinear approximate graph unlearning method based on GSTs. Our second
contribution is a theoretical analysis of the computational complexity of the
proposed unlearning mechanism, which is hard to replicate for deep neural
networks. Our third contribution are extensive simulation results which show
that, compared to complete retraining of GNNs after each removal request, the
new GST-based approach offers, on average, a 10.38x speed-up and leads to a
2.6% increase in test accuracy during unlearning of 90 out of 100 training
graphs from the IMDB dataset (10% training ratio). Our implementation is
available online at https://doi.org/10.5281/zenodo.7613150.
| [
{
"version": "v1",
"created": "Sun, 6 Nov 2022 20:46:50 GMT"
},
{
"version": "v2",
"created": "Sat, 1 Jul 2023 04:05:49 GMT"
}
] | 2023-07-04T00:00:00 | [
[
"Pan",
"Chao",
""
],
[
"Chien",
"Eli",
""
],
[
"Milenkovic",
"Olgica",
""
]
] | TITLE: Unlearning Graph Classifiers with Limited Data Resources
ABSTRACT: As the demand for user privacy grows, controlled data removal (machine
unlearning) is becoming an important feature of machine learning models for
data-sensitive Web applications such as social networks and recommender
systems. Nevertheless, at this point it is still largely unknown how to perform
efficient machine unlearning of graph neural networks (GNNs); this is
especially the case when the number of training samples is small, in which case
unlearning can seriously compromise the performance of the model. To address
this issue, we initiate the study of unlearning the Graph Scattering Transform
(GST), a mathematical framework that is efficient, provably stable under
feature or graph topology perturbations, and offers graph classification
performance comparable to that of GNNs. Our main contribution is the first
known nonlinear approximate graph unlearning method based on GSTs. Our second
contribution is a theoretical analysis of the computational complexity of the
proposed unlearning mechanism, which is hard to replicate for deep neural
networks. Our third contribution are extensive simulation results which show
that, compared to complete retraining of GNNs after each removal request, the
new GST-based approach offers, on average, a 10.38x speed-up and leads to a
2.6% increase in test accuracy during unlearning of 90 out of 100 training
graphs from the IMDB dataset (10% training ratio). Our implementation is
available online at https://doi.org/10.5281/zenodo.7613150.
| no_new_dataset | 0.70878 |
2211.05039 | Jannik Kossen | Jannik Kossen, C\u{a}t\u{a}lina Cangea, Eszter V\'ertes, Andrew
Jaegle, Viorica Patraucean, Ira Ktena, Nenad Tomasev, Danielle Belgrave | Active Acquisition for Multimodal Temporal Data: A Challenging
Decision-Making Task | Published in Transactions on Machine Learning Research. Previous
version accepted to Foundation Models for Decision Making Workshop at NeurIPS
2022 | null | null | null | cs.LG cs.AI cs.CV stat.ML | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We introduce a challenging decision-making task that we call active
acquisition for multimodal temporal data (A2MT). In many real-world scenarios,
input features are not readily available at test time and must instead be
acquired at significant cost. With A2MT, we aim to learn agents that actively
select which modalities of an input to acquire, trading off acquisition cost
and predictive performance. A2MT extends a previous task called active feature
acquisition to temporal decision making about high-dimensional inputs. We
propose a method based on the Perceiver IO architecture to address A2MT in
practice. Our agents are able to solve a novel synthetic scenario requiring
practically relevant cross-modal reasoning skills. On two large-scale,
real-world datasets, Kinetics-700 and AudioSet, our agents successfully learn
cost-reactive acquisition behavior. However, an ablation reveals they are
unable to learn adaptive acquisition strategies, emphasizing the difficulty of
the task even for state-of-the-art models. Applications of A2MT may be
impactful in domains like medicine, robotics, or finance, where modalities
differ in acquisition cost and informativeness.
| [
{
"version": "v1",
"created": "Wed, 9 Nov 2022 17:16:36 GMT"
},
{
"version": "v2",
"created": "Mon, 3 Jul 2023 14:47:18 GMT"
}
] | 2023-07-04T00:00:00 | [
[
"Kossen",
"Jannik",
""
],
[
"Cangea",
"Cătălina",
""
],
[
"Vértes",
"Eszter",
""
],
[
"Jaegle",
"Andrew",
""
],
[
"Patraucean",
"Viorica",
""
],
[
"Ktena",
"Ira",
""
],
[
"Tomasev",
"Nenad",
""
],
[
"Belgrave",
"Danielle",
""
]
] | TITLE: Active Acquisition for Multimodal Temporal Data: A Challenging
Decision-Making Task
ABSTRACT: We introduce a challenging decision-making task that we call active
acquisition for multimodal temporal data (A2MT). In many real-world scenarios,
input features are not readily available at test time and must instead be
acquired at significant cost. With A2MT, we aim to learn agents that actively
select which modalities of an input to acquire, trading off acquisition cost
and predictive performance. A2MT extends a previous task called active feature
acquisition to temporal decision making about high-dimensional inputs. We
propose a method based on the Perceiver IO architecture to address A2MT in
practice. Our agents are able to solve a novel synthetic scenario requiring
practically relevant cross-modal reasoning skills. On two large-scale,
real-world datasets, Kinetics-700 and AudioSet, our agents successfully learn
cost-reactive acquisition behavior. However, an ablation reveals they are
unable to learn adaptive acquisition strategies, emphasizing the difficulty of
the task even for state-of-the-art models. Applications of A2MT may be
impactful in domains like medicine, robotics, or finance, where modalities
differ in acquisition cost and informativeness.
| no_new_dataset | 0.706013 |
2211.07467 | Leonard Bauersfeld | Leonard Bauersfeld and Angel Romero and Manasi Muglikar and Davide
Scaramuzza | Cracking Double-Blind Review: Authorship Attribution with Deep Learning | 13 pages + 3 pages references | PLOS ONE 18(6): e0287611 (2023) | 10.1371/journal.pone.0287611 | null | cs.CL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Double-blind peer review is considered a pillar of academic research because
it is perceived to ensure a fair, unbiased, and fact-centered scientific
discussion. Yet, experienced researchers can often correctly guess from which
research group an anonymous submission originates, biasing the peer-review
process. In this work, we present a transformer-based, neural-network
architecture that only uses the text content and the author names in the
bibliography to attribute an anonymous manuscript to an author. To train and
evaluate our method, we created the largest authorship identification dataset
to date. It leverages all research papers publicly available on arXiv amounting
to over 2 million manuscripts. In arXiv-subsets with up to 2,000 different
authors, our method achieves an unprecedented authorship attribution accuracy,
where up to 73% of papers are attributed correctly. We present a scaling
analysis to highlight the applicability of the proposed method to even larger
datasets when sufficient compute capabilities are more widely available to the
academic community. Furthermore, we analyze the attribution accuracy in
settings where the goal is to identify all authors of an anonymous manuscript.
Thanks to our method, we are not only able to predict the author of an
anonymous work, but we also provide empirical evidence of the key aspects that
make a paper attributable. We have open-sourced the necessary tools to
reproduce our experiments.
| [
{
"version": "v1",
"created": "Mon, 14 Nov 2022 15:50:24 GMT"
},
{
"version": "v2",
"created": "Wed, 30 Nov 2022 09:31:03 GMT"
},
{
"version": "v3",
"created": "Mon, 3 Jul 2023 12:49:54 GMT"
}
] | 2023-07-04T00:00:00 | [
[
"Bauersfeld",
"Leonard",
""
],
[
"Romero",
"Angel",
""
],
[
"Muglikar",
"Manasi",
""
],
[
"Scaramuzza",
"Davide",
""
]
] | TITLE: Cracking Double-Blind Review: Authorship Attribution with Deep Learning
ABSTRACT: Double-blind peer review is considered a pillar of academic research because
it is perceived to ensure a fair, unbiased, and fact-centered scientific
discussion. Yet, experienced researchers can often correctly guess from which
research group an anonymous submission originates, biasing the peer-review
process. In this work, we present a transformer-based, neural-network
architecture that only uses the text content and the author names in the
bibliography to attribute an anonymous manuscript to an author. To train and
evaluate our method, we created the largest authorship identification dataset
to date. It leverages all research papers publicly available on arXiv amounting
to over 2 million manuscripts. In arXiv-subsets with up to 2,000 different
authors, our method achieves an unprecedented authorship attribution accuracy,
where up to 73% of papers are attributed correctly. We present a scaling
analysis to highlight the applicability of the proposed method to even larger
datasets when sufficient compute capabilities are more widely available to the
academic community. Furthermore, we analyze the attribution accuracy in
settings where the goal is to identify all authors of an anonymous manuscript.
Thanks to our method, we are not only able to predict the author of an
anonymous work, but we also provide empirical evidence of the key aspects that
make a paper attributable. We have open-sourced the necessary tools to
reproduce our experiments.
| new_dataset | 0.71625 |
2211.11736 | Ted Xiao | Ted Xiao and Harris Chan and Pierre Sermanet and Ayzaan Wahid and
Anthony Brohan and Karol Hausman and Sergey Levine and Jonathan Tompson | Robotic Skill Acquisition via Instruction Augmentation with
Vision-Language Models | Published as a conference paper at RSS 2023 | null | null | null | cs.RO cs.AI cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In recent years, much progress has been made in learning robotic manipulation
policies that follow natural language instructions. Such methods typically
learn from corpora of robot-language data that was either collected with
specific tasks in mind or expensively re-labelled by humans with rich language
descriptions in hindsight. Recently, large-scale pretrained vision-language
models (VLMs) like CLIP or ViLD have been applied to robotics for learning
representations and scene descriptors. Can these pretrained models serve as
automatic labelers for robot data, effectively importing Internet-scale
knowledge into existing datasets to make them useful even for tasks that are
not reflected in their ground truth annotations? To accomplish this, we
introduce Data-driven Instruction Augmentation for Language-conditioned control
(DIAL): we utilize semi-supervised language labels leveraging the semantic
understanding of CLIP to propagate knowledge onto large datasets of unlabelled
demonstration data and then train language-conditioned policies on the
augmented datasets. This method enables cheaper acquisition of useful language
descriptions compared to expensive human labels, allowing for more efficient
label coverage of large-scale datasets. We apply DIAL to a challenging
real-world robotic manipulation domain where 96.5% of the 80,000 demonstrations
do not contain crowd-sourced language annotations. DIAL enables imitation
learning policies to acquire new capabilities and generalize to 60 novel
instructions unseen in the original dataset.
| [
{
"version": "v1",
"created": "Mon, 21 Nov 2022 18:56:00 GMT"
},
{
"version": "v2",
"created": "Tue, 22 Nov 2022 07:31:59 GMT"
},
{
"version": "v3",
"created": "Sat, 1 Jul 2023 05:38:52 GMT"
}
] | 2023-07-04T00:00:00 | [
[
"Xiao",
"Ted",
""
],
[
"Chan",
"Harris",
""
],
[
"Sermanet",
"Pierre",
""
],
[
"Wahid",
"Ayzaan",
""
],
[
"Brohan",
"Anthony",
""
],
[
"Hausman",
"Karol",
""
],
[
"Levine",
"Sergey",
""
],
[
"Tompson",
"Jonathan",
""
]
] | TITLE: Robotic Skill Acquisition via Instruction Augmentation with
Vision-Language Models
ABSTRACT: In recent years, much progress has been made in learning robotic manipulation
policies that follow natural language instructions. Such methods typically
learn from corpora of robot-language data that was either collected with
specific tasks in mind or expensively re-labelled by humans with rich language
descriptions in hindsight. Recently, large-scale pretrained vision-language
models (VLMs) like CLIP or ViLD have been applied to robotics for learning
representations and scene descriptors. Can these pretrained models serve as
automatic labelers for robot data, effectively importing Internet-scale
knowledge into existing datasets to make them useful even for tasks that are
not reflected in their ground truth annotations? To accomplish this, we
introduce Data-driven Instruction Augmentation for Language-conditioned control
(DIAL): we utilize semi-supervised language labels leveraging the semantic
understanding of CLIP to propagate knowledge onto large datasets of unlabelled
demonstration data and then train language-conditioned policies on the
augmented datasets. This method enables cheaper acquisition of useful language
descriptions compared to expensive human labels, allowing for more efficient
label coverage of large-scale datasets. We apply DIAL to a challenging
real-world robotic manipulation domain where 96.5% of the 80,000 demonstrations
do not contain crowd-sourced language annotations. DIAL enables imitation
learning policies to acquire new capabilities and generalize to 60 novel
instructions unseen in the original dataset.
| no_new_dataset | 0.710842 |
2211.15875 | Jaehyun Choi | Gyojin Han, Jaehyun Choi, Hyeong Gwon Hong, Junmo Kim | Data Poisoning Attack Aiming the Vulnerability of Continual Learning | ICIP 2023 (NeurIPS 2022 ML Safety Workshop accepted paper) | null | null | null | cs.LG cs.CR cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Generally, regularization-based continual learning models limit access to the
previous task data to imitate the real-world constraints related to memory and
privacy. However, this introduces a problem in these models by not being able
to track the performance on each task. In essence, current continual learning
methods are susceptible to attacks on previous tasks. We demonstrate the
vulnerability of regularization-based continual learning methods by presenting
a simple task-specific data poisoning attack that can be used in the learning
process of a new task. Training data generated by the proposed attack causes
performance degradation on a specific task targeted by the attacker. We
experiment with the attack on the two representative regularization-based
continual learning methods, Elastic Weight Consolidation (EWC) and Synaptic
Intelligence (SI), trained with variants of MNIST dataset. The experiment
results justify the vulnerability proposed in this paper and demonstrate the
importance of developing continual learning models that are robust to
adversarial attacks.
| [
{
"version": "v1",
"created": "Tue, 29 Nov 2022 02:28:05 GMT"
},
{
"version": "v2",
"created": "Mon, 3 Jul 2023 09:50:12 GMT"
}
] | 2023-07-04T00:00:00 | [
[
"Han",
"Gyojin",
""
],
[
"Choi",
"Jaehyun",
""
],
[
"Hong",
"Hyeong Gwon",
""
],
[
"Kim",
"Junmo",
""
]
] | TITLE: Data Poisoning Attack Aiming the Vulnerability of Continual Learning
ABSTRACT: Generally, regularization-based continual learning models limit access to the
previous task data to imitate the real-world constraints related to memory and
privacy. However, this introduces a problem in these models by not being able
to track the performance on each task. In essence, current continual learning
methods are susceptible to attacks on previous tasks. We demonstrate the
vulnerability of regularization-based continual learning methods by presenting
a simple task-specific data poisoning attack that can be used in the learning
process of a new task. Training data generated by the proposed attack causes
performance degradation on a specific task targeted by the attacker. We
experiment with the attack on the two representative regularization-based
continual learning methods, Elastic Weight Consolidation (EWC) and Synaptic
Intelligence (SI), trained with variants of MNIST dataset. The experiment
results justify the vulnerability proposed in this paper and demonstrate the
importance of developing continual learning models that are robust to
adversarial attacks.
| no_new_dataset | 0.709799 |
2211.16779 | Yunzhe Wu | Zizhang Wu, Yunzhe Wu, Jian Pu, Xianzhi Li and Xiaoquan Wang | Attention-Based Depth Distillation with 3D-Aware Positional Encoding for
Monocular 3D Object Detection | Accepted by AAAI2023 | null | 10.1609/aaai.v37i3.25391 | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Monocular 3D object detection is a low-cost but challenging task, as it
requires generating accurate 3D localization solely from a single image input.
Recent developed depth-assisted methods show promising results by using
explicit depth maps as intermediate features, which are either precomputed by
monocular depth estimation networks or jointly evaluated with 3D object
detection. However, inevitable errors from estimated depth priors may lead to
misaligned semantic information and 3D localization, hence resulting in feature
smearing and suboptimal predictions. To mitigate this issue, we propose ADD, an
Attention-based Depth knowledge Distillation framework with 3D-aware positional
encoding. Unlike previous knowledge distillation frameworks that adopt stereo-
or LiDAR-based teachers, we build up our teacher with identical architecture as
the student but with extra ground-truth depth as input. Credit to our teacher
design, our framework is seamless, domain-gap free, easily implementable, and
is compatible with object-wise ground-truth depth. Specifically, we leverage
intermediate features and responses for knowledge distillation. Considering
long-range 3D dependencies, we propose \emph{3D-aware self-attention} and
\emph{target-aware cross-attention} modules for student adaptation. Extensive
experiments are performed to verify the effectiveness of our framework on the
challenging KITTI 3D object detection benchmark. We implement our framework on
three representative monocular detectors, and we achieve state-of-the-art
performance with no additional inference computational cost relative to
baseline models. Our code is available at https://github.com/rockywind/ADD.
| [
{
"version": "v1",
"created": "Wed, 30 Nov 2022 06:39:25 GMT"
},
{
"version": "v2",
"created": "Mon, 3 Jul 2023 08:07:45 GMT"
}
] | 2023-07-04T00:00:00 | [
[
"Wu",
"Zizhang",
""
],
[
"Wu",
"Yunzhe",
""
],
[
"Pu",
"Jian",
""
],
[
"Li",
"Xianzhi",
""
],
[
"Wang",
"Xiaoquan",
""
]
] | TITLE: Attention-Based Depth Distillation with 3D-Aware Positional Encoding for
Monocular 3D Object Detection
ABSTRACT: Monocular 3D object detection is a low-cost but challenging task, as it
requires generating accurate 3D localization solely from a single image input.
Recent developed depth-assisted methods show promising results by using
explicit depth maps as intermediate features, which are either precomputed by
monocular depth estimation networks or jointly evaluated with 3D object
detection. However, inevitable errors from estimated depth priors may lead to
misaligned semantic information and 3D localization, hence resulting in feature
smearing and suboptimal predictions. To mitigate this issue, we propose ADD, an
Attention-based Depth knowledge Distillation framework with 3D-aware positional
encoding. Unlike previous knowledge distillation frameworks that adopt stereo-
or LiDAR-based teachers, we build up our teacher with identical architecture as
the student but with extra ground-truth depth as input. Credit to our teacher
design, our framework is seamless, domain-gap free, easily implementable, and
is compatible with object-wise ground-truth depth. Specifically, we leverage
intermediate features and responses for knowledge distillation. Considering
long-range 3D dependencies, we propose \emph{3D-aware self-attention} and
\emph{target-aware cross-attention} modules for student adaptation. Extensive
experiments are performed to verify the effectiveness of our framework on the
challenging KITTI 3D object detection benchmark. We implement our framework on
three representative monocular detectors, and we achieve state-of-the-art
performance with no additional inference computational cost relative to
baseline models. Our code is available at https://github.com/rockywind/ADD.
| no_new_dataset | 0.708811 |
2212.10180 | Ananya B Sai | Ananya B. Sai, Vignesh Nagarajan, Tanay Dixit, Raj Dabre, Anoop
Kunchukuttan, Pratyush Kumar, Mitesh M. Khapra | IndicMT Eval: A Dataset to Meta-Evaluate Machine Translation metrics for
Indian Languages | ACL 2023 long paper | null | null | null | cs.CL | http://creativecommons.org/licenses/by/4.0/ | The rapid growth of machine translation (MT) systems has necessitated
comprehensive studies to meta-evaluate evaluation metrics being used, which
enables a better selection of metrics that best reflect MT quality.
Unfortunately, most of the research focuses on high-resource languages, mainly
English, the observations for which may not always apply to other languages.
Indian languages, having over a billion speakers, are linguistically different
from English, and to date, there has not been a systematic study of evaluating
MT systems from English into Indian languages. In this paper, we fill this gap
by creating an MQM dataset consisting of 7000 fine-grained annotations,
spanning 5 Indian languages and 7 MT systems, and use it to establish
correlations between annotator scores and scores obtained using existing
automatic metrics. Our results show that pre-trained metrics, such as COMET,
have the highest correlations with annotator scores. Additionally, we find that
the metrics do not adequately capture fluency-based errors in Indian languages,
and there is a need to develop metrics focused on Indian languages. We hope
that our dataset and analysis will help promote further research in this area.
| [
{
"version": "v1",
"created": "Tue, 20 Dec 2022 11:37:22 GMT"
},
{
"version": "v2",
"created": "Mon, 3 Jul 2023 14:26:38 GMT"
}
] | 2023-07-04T00:00:00 | [
[
"Sai",
"Ananya B.",
""
],
[
"Nagarajan",
"Vignesh",
""
],
[
"Dixit",
"Tanay",
""
],
[
"Dabre",
"Raj",
""
],
[
"Kunchukuttan",
"Anoop",
""
],
[
"Kumar",
"Pratyush",
""
],
[
"Khapra",
"Mitesh M.",
""
]
] | TITLE: IndicMT Eval: A Dataset to Meta-Evaluate Machine Translation metrics for
Indian Languages
ABSTRACT: The rapid growth of machine translation (MT) systems has necessitated
comprehensive studies to meta-evaluate evaluation metrics being used, which
enables a better selection of metrics that best reflect MT quality.
Unfortunately, most of the research focuses on high-resource languages, mainly
English, the observations for which may not always apply to other languages.
Indian languages, having over a billion speakers, are linguistically different
from English, and to date, there has not been a systematic study of evaluating
MT systems from English into Indian languages. In this paper, we fill this gap
by creating an MQM dataset consisting of 7000 fine-grained annotations,
spanning 5 Indian languages and 7 MT systems, and use it to establish
correlations between annotator scores and scores obtained using existing
automatic metrics. Our results show that pre-trained metrics, such as COMET,
have the highest correlations with annotator scores. Additionally, we find that
the metrics do not adequately capture fluency-based errors in Indian languages,
and there is a need to develop metrics focused on Indian languages. We hope
that our dataset and analysis will help promote further research in this area.
| new_dataset | 0.717804 |
2212.10511 | Alex Mallen | Alex Mallen, Akari Asai, Victor Zhong, Rajarshi Das, Daniel Khashabi,
Hannaneh Hajishirzi | When Not to Trust Language Models: Investigating Effectiveness of
Parametric and Non-Parametric Memories | ACL 2023; Code and data available at
https://github.com/AlexTMallen/adaptive-retrieval | null | null | null | cs.CL cs.AI cs.LG | http://creativecommons.org/licenses/by/4.0/ | Despite their impressive performance on diverse tasks, large language models
(LMs) still struggle with tasks requiring rich world knowledge, implying the
limitations of relying solely on their parameters to encode a wealth of world
knowledge. This paper aims to understand LMs' strengths and limitations in
memorizing factual knowledge, by conducting large-scale knowledge probing
experiments of 10 models and 4 augmentation methods on PopQA, our new
open-domain QA dataset with 14k questions. We find that LMs struggle with less
popular factual knowledge, and that scaling fails to appreciably improve
memorization of factual knowledge in the long tail. We then show that
retrieval-augmented LMs largely outperform orders of magnitude larger LMs,
while unassisted LMs remain competitive in questions about high-popularity
entities. Based on those findings, we devise a simple, yet effective, method
for powerful and efficient retrieval-augmented LMs, which retrieves
non-parametric memories only when necessary. Experimental results show that
this significantly improves models' performance while reducing the inference
costs.
| [
{
"version": "v1",
"created": "Tue, 20 Dec 2022 18:30:15 GMT"
},
{
"version": "v2",
"created": "Sun, 28 May 2023 18:45:38 GMT"
},
{
"version": "v3",
"created": "Tue, 30 May 2023 17:27:40 GMT"
},
{
"version": "v4",
"created": "Sun, 2 Jul 2023 07:21:59 GMT"
}
] | 2023-07-04T00:00:00 | [
[
"Mallen",
"Alex",
""
],
[
"Asai",
"Akari",
""
],
[
"Zhong",
"Victor",
""
],
[
"Das",
"Rajarshi",
""
],
[
"Khashabi",
"Daniel",
""
],
[
"Hajishirzi",
"Hannaneh",
""
]
] | TITLE: When Not to Trust Language Models: Investigating Effectiveness of
Parametric and Non-Parametric Memories
ABSTRACT: Despite their impressive performance on diverse tasks, large language models
(LMs) still struggle with tasks requiring rich world knowledge, implying the
limitations of relying solely on their parameters to encode a wealth of world
knowledge. This paper aims to understand LMs' strengths and limitations in
memorizing factual knowledge, by conducting large-scale knowledge probing
experiments of 10 models and 4 augmentation methods on PopQA, our new
open-domain QA dataset with 14k questions. We find that LMs struggle with less
popular factual knowledge, and that scaling fails to appreciably improve
memorization of factual knowledge in the long tail. We then show that
retrieval-augmented LMs largely outperform orders of magnitude larger LMs,
while unassisted LMs remain competitive in questions about high-popularity
entities. Based on those findings, we devise a simple, yet effective, method
for powerful and efficient retrieval-augmented LMs, which retrieves
non-parametric memories only when necessary. Experimental results show that
this significantly improves models' performance while reducing the inference
costs.
| new_dataset | 0.715387 |
2301.02200 | Daniel Bogdoll | Daniel Bogdoll, Jonas Hendl, Felix Schreyer, Nishanth Gowda, Michael
F\"arber, J. Marius Z\"ollner | Impact, Attention, Influence: Early Assessment of Autonomous Driving
Datasets | Daniel Bogdoll and Jonas Hendl contributed equally. Accepted for
publication at ICCRE 2023 | null | 10.1109/ICCRE57112.2023.10155607 | null | cs.DL cs.RO | http://creativecommons.org/licenses/by/4.0/ | Autonomous Driving (AD), the area of robotics with the greatest potential
impact on society, has gained a lot of momentum in the last decade. As a result
of this, the number of datasets in AD has increased rapidly. Creators and users
of datasets can benefit from a better understanding of developments in the
field. While scientometric analysis has been conducted in other fields, it
rarely revolves around datasets. Thus, the impact, attention, and influence of
datasets on autonomous driving remains a rarely investigated field. In this
work, we provide a scientometric analysis for over 200 datasets in AD. We
perform a rigorous evaluation of relations between available metadata and
citation counts based on linear regression. Subsequently, we propose an
Influence Score to assess a dataset already early on without the need for a
track-record of citations, which is only available with a certain delay.
| [
{
"version": "v1",
"created": "Thu, 5 Jan 2023 18:07:15 GMT"
},
{
"version": "v2",
"created": "Fri, 31 Mar 2023 12:40:37 GMT"
}
] | 2023-07-04T00:00:00 | [
[
"Bogdoll",
"Daniel",
""
],
[
"Hendl",
"Jonas",
""
],
[
"Schreyer",
"Felix",
""
],
[
"Gowda",
"Nishanth",
""
],
[
"Färber",
"Michael",
""
],
[
"Zöllner",
"J. Marius",
""
]
] | TITLE: Impact, Attention, Influence: Early Assessment of Autonomous Driving
Datasets
ABSTRACT: Autonomous Driving (AD), the area of robotics with the greatest potential
impact on society, has gained a lot of momentum in the last decade. As a result
of this, the number of datasets in AD has increased rapidly. Creators and users
of datasets can benefit from a better understanding of developments in the
field. While scientometric analysis has been conducted in other fields, it
rarely revolves around datasets. Thus, the impact, attention, and influence of
datasets on autonomous driving remains a rarely investigated field. In this
work, we provide a scientometric analysis for over 200 datasets in AD. We
perform a rigorous evaluation of relations between available metadata and
citation counts based on linear regression. Subsequently, we propose an
Influence Score to assess a dataset already early on without the need for a
track-record of citations, which is only available with a certain delay.
| no_new_dataset | 0.71262 |
2301.04318 | Zhihao Wu | Shiping Wang, Zhihao Wu, Yuhong Chen, Yong Chen | Beyond Graph Convolutional Network: An Interpretable
Regularizer-centered Optimization Framework | 15 pages, 12 figures | null | 10.1609/aaai.v37i4.25593 | null | cs.LG cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Graph convolutional networks (GCNs) have been attracting widespread
attentions due to their encouraging performance and powerful generalizations.
However, few work provide a general view to interpret various GCNs and guide
GCNs' designs. In this paper, by revisiting the original GCN, we induce an
interpretable regularizer-centerd optimization framework, in which by building
appropriate regularizers we can interpret most GCNs, such as APPNP, JKNet,
DAGNN, and GNN-LF/HF. Further, under the proposed framework, we devise a
dual-regularizer graph convolutional network (dubbed tsGCN) to capture
topological and semantic structures from graph data. Since the derived learning
rule for tsGCN contains an inverse of a large matrix and thus is
time-consuming, we leverage the Woodbury matrix identity and low-rank
approximation tricks to successfully decrease the high computational complexity
of computing infinite-order graph convolutions. Extensive experiments on eight
public datasets demonstrate that tsGCN achieves superior performance against
quite a few state-of-the-art competitors w.r.t. classification tasks.
| [
{
"version": "v1",
"created": "Wed, 11 Jan 2023 05:51:33 GMT"
}
] | 2023-07-04T00:00:00 | [
[
"Wang",
"Shiping",
""
],
[
"Wu",
"Zhihao",
""
],
[
"Chen",
"Yuhong",
""
],
[
"Chen",
"Yong",
""
]
] | TITLE: Beyond Graph Convolutional Network: An Interpretable
Regularizer-centered Optimization Framework
ABSTRACT: Graph convolutional networks (GCNs) have been attracting widespread
attentions due to their encouraging performance and powerful generalizations.
However, few work provide a general view to interpret various GCNs and guide
GCNs' designs. In this paper, by revisiting the original GCN, we induce an
interpretable regularizer-centerd optimization framework, in which by building
appropriate regularizers we can interpret most GCNs, such as APPNP, JKNet,
DAGNN, and GNN-LF/HF. Further, under the proposed framework, we devise a
dual-regularizer graph convolutional network (dubbed tsGCN) to capture
topological and semantic structures from graph data. Since the derived learning
rule for tsGCN contains an inverse of a large matrix and thus is
time-consuming, we leverage the Woodbury matrix identity and low-rank
approximation tricks to successfully decrease the high computational complexity
of computing infinite-order graph convolutions. Extensive experiments on eight
public datasets demonstrate that tsGCN achieves superior performance against
quite a few state-of-the-art competitors w.r.t. classification tasks.
| no_new_dataset | 0.709435 |
2301.06264 | DeHua Liu | Dehua Liu, Selasi Kwashie, Yidi Zhang, Guangtong Zhou, Michael Bewong,
Xiaoying Wu, Xi Guo, Keqing He, Zaiwen Feng | An Efficient Approach for Discovering Graph Entity Dependencies (GEDs) | null | null | null | null | cs.DB | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Graph entity dependencies (GEDs) are novel graph constraints, unifying keys
and functional dependencies, for property graphs. They have been found useful
in many real-world data quality and data management tasks, including fact
checking on social media networks and entity resolution. In this paper, we
study the discovery problem of GEDs -- finding a minimal cover of valid GEDs in
a given graph data. We formalise the problem, and propose an effective and
efficient approach to overcome major bottlenecks in GED discovery. In
particular, we leverage existing graph partitioning algorithms to enable fast
GED-scope discovery, and employ effective pruning strategies over the
prohibitively large space of candidate dependencies. Furthermore, we define an
interestingness measure for GEDs based on the minimum description length
principle, to score and rank the mined cover set of GEDs. Finally, we
demonstrate the scalability and effectiveness of our GED discovery approach
through extensive experiments on real-world benchmark graph data sets; and
present the usefulness of the discovered rules in different downstream data
quality management applications.
| [
{
"version": "v1",
"created": "Mon, 16 Jan 2023 05:21:34 GMT"
},
{
"version": "v2",
"created": "Sat, 1 Jul 2023 01:07:14 GMT"
}
] | 2023-07-04T00:00:00 | [
[
"Liu",
"Dehua",
""
],
[
"Kwashie",
"Selasi",
""
],
[
"Zhang",
"Yidi",
""
],
[
"Zhou",
"Guangtong",
""
],
[
"Bewong",
"Michael",
""
],
[
"Wu",
"Xiaoying",
""
],
[
"Guo",
"Xi",
""
],
[
"He",
"Keqing",
""
],
[
"Feng",
"Zaiwen",
""
]
] | TITLE: An Efficient Approach for Discovering Graph Entity Dependencies (GEDs)
ABSTRACT: Graph entity dependencies (GEDs) are novel graph constraints, unifying keys
and functional dependencies, for property graphs. They have been found useful
in many real-world data quality and data management tasks, including fact
checking on social media networks and entity resolution. In this paper, we
study the discovery problem of GEDs -- finding a minimal cover of valid GEDs in
a given graph data. We formalise the problem, and propose an effective and
efficient approach to overcome major bottlenecks in GED discovery. In
particular, we leverage existing graph partitioning algorithms to enable fast
GED-scope discovery, and employ effective pruning strategies over the
prohibitively large space of candidate dependencies. Furthermore, we define an
interestingness measure for GEDs based on the minimum description length
principle, to score and rank the mined cover set of GEDs. Finally, we
demonstrate the scalability and effectiveness of our GED discovery approach
through extensive experiments on real-world benchmark graph data sets; and
present the usefulness of the discovered rules in different downstream data
quality management applications.
| no_new_dataset | 0.7114 |
2302.09543 | Antonio Briola | Antonio Briola and Tomaso Aste | Topological Feature Selection | Accepted at the 2nd Annual Workshop on Topology, Algebra, and
Geometry in Machine Learning (TAG-ML) at the 40th International Conference on
Machine Learning, Honolulu, Hawaii, USA. 2023. 23 pages, 2 figures, 13 tables | null | null | null | cs.LG cs.AI | http://creativecommons.org/licenses/by-nc-nd/4.0/ | In this paper, we introduce a novel unsupervised, graph-based filter feature
selection technique which exploits the power of topologically constrained
network representations. We model dependency structures among features using a
family of chordal graphs (the Triangulated Maximally Filtered Graph), and we
maximise the likelihood of features' relevance by studying their relative
position inside the network. Such an approach presents three aspects that are
particularly satisfactory compared to its alternatives: (i) it is highly
tunable and easily adaptable to the nature of input data; (ii) it is fully
explainable, maintaining, at the same time, a remarkable level of simplicity;
(iii) it is computationally cheaper compared to its alternatives. We test our
algorithm on 16 benchmark datasets from different applicative domains showing
that it outperforms or matches the current state-of-the-art under heterogeneous
evaluation conditions.
| [
{
"version": "v1",
"created": "Sun, 19 Feb 2023 11:29:15 GMT"
},
{
"version": "v2",
"created": "Wed, 5 Apr 2023 06:36:25 GMT"
},
{
"version": "v3",
"created": "Sat, 1 Jul 2023 08:08:54 GMT"
}
] | 2023-07-04T00:00:00 | [
[
"Briola",
"Antonio",
""
],
[
"Aste",
"Tomaso",
""
]
] | TITLE: Topological Feature Selection
ABSTRACT: In this paper, we introduce a novel unsupervised, graph-based filter feature
selection technique which exploits the power of topologically constrained
network representations. We model dependency structures among features using a
family of chordal graphs (the Triangulated Maximally Filtered Graph), and we
maximise the likelihood of features' relevance by studying their relative
position inside the network. Such an approach presents three aspects that are
particularly satisfactory compared to its alternatives: (i) it is highly
tunable and easily adaptable to the nature of input data; (ii) it is fully
explainable, maintaining, at the same time, a remarkable level of simplicity;
(iii) it is computationally cheaper compared to its alternatives. We test our
algorithm on 16 benchmark datasets from different applicative domains showing
that it outperforms or matches the current state-of-the-art under heterogeneous
evaluation conditions.
| no_new_dataset | 0.709239 |
2302.10258 | Beatrice Bevilacqua | Beatrice Bevilacqua, Kyriacos Nikiforou, Borja Ibarz, Ioana Bica,
Michela Paganini, Charles Blundell, Jovana Mitrovic, Petar Veli\v{c}kovi\'c | Neural Algorithmic Reasoning with Causal Regularisation | ICML 2023, Camera Ready; 17 pages, 7 figures | null | null | null | cs.LG cs.AI stat.ME | http://creativecommons.org/licenses/by/4.0/ | Recent work on neural algorithmic reasoning has investigated the reasoning
capabilities of neural networks, effectively demonstrating they can learn to
execute classical algorithms on unseen data coming from the train distribution.
However, the performance of existing neural reasoners significantly degrades on
out-of-distribution (OOD) test data, where inputs have larger sizes. In this
work, we make an important observation: there are many different inputs for
which an algorithm will perform certain intermediate computations identically.
This insight allows us to develop data augmentation procedures that, given an
algorithm's intermediate trajectory, produce inputs for which the target
algorithm would have exactly the same next trajectory step. We ensure
invariance in the next-step prediction across such inputs, by employing a
self-supervised objective derived by our observation, formalised in a causal
graph. We prove that the resulting method, which we call Hint-ReLIC, improves
the OOD generalisation capabilities of the reasoner. We evaluate our method on
the CLRS algorithmic reasoning benchmark, where we show up to 3$\times$
improvements on the OOD test data.
| [
{
"version": "v1",
"created": "Mon, 20 Feb 2023 19:41:15 GMT"
},
{
"version": "v2",
"created": "Mon, 3 Jul 2023 17:08:05 GMT"
}
] | 2023-07-04T00:00:00 | [
[
"Bevilacqua",
"Beatrice",
""
],
[
"Nikiforou",
"Kyriacos",
""
],
[
"Ibarz",
"Borja",
""
],
[
"Bica",
"Ioana",
""
],
[
"Paganini",
"Michela",
""
],
[
"Blundell",
"Charles",
""
],
[
"Mitrovic",
"Jovana",
""
],
[
"Veličković",
"Petar",
""
]
] | TITLE: Neural Algorithmic Reasoning with Causal Regularisation
ABSTRACT: Recent work on neural algorithmic reasoning has investigated the reasoning
capabilities of neural networks, effectively demonstrating they can learn to
execute classical algorithms on unseen data coming from the train distribution.
However, the performance of existing neural reasoners significantly degrades on
out-of-distribution (OOD) test data, where inputs have larger sizes. In this
work, we make an important observation: there are many different inputs for
which an algorithm will perform certain intermediate computations identically.
This insight allows us to develop data augmentation procedures that, given an
algorithm's intermediate trajectory, produce inputs for which the target
algorithm would have exactly the same next trajectory step. We ensure
invariance in the next-step prediction across such inputs, by employing a
self-supervised objective derived by our observation, formalised in a causal
graph. We prove that the resulting method, which we call Hint-ReLIC, improves
the OOD generalisation capabilities of the reasoner. We evaluate our method on
the CLRS algorithmic reasoning benchmark, where we show up to 3$\times$
improvements on the OOD test data.
| no_new_dataset | 0.70779 |
2302.14260 | Sungbin Shin | Sungbin Shin, Yohan Jo, Sungsoo Ahn, Namhoon Lee | A Closer Look at the Intervention Procedure of Concept Bottleneck Models | ICML 2023 | null | null | null | cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Concept bottleneck models (CBMs) are a class of interpretable neural network
models that predict the target response of a given input based on its
high-level concepts. Unlike the standard end-to-end models, CBMs enable domain
experts to intervene on the predicted concepts and rectify any mistakes at test
time, so that more accurate task predictions can be made at the end. While such
intervenability provides a powerful avenue of control, many aspects of the
intervention procedure remain rather unexplored. In this work, we develop
various ways of selecting intervening concepts to improve the intervention
effectiveness and conduct an array of in-depth analyses as to how they evolve
under different circumstances. Specifically, we find that an informed
intervention strategy can reduce the task error more than ten times compared to
the current baseline under the same amount of intervention counts in realistic
settings, and yet, this can vary quite significantly when taking into account
different intervention granularity. We verify our findings through
comprehensive evaluations, not only on the standard real datasets, but also on
synthetic datasets that we generate based on a set of different causal graphs.
We further discover some major pitfalls of the current practices which, without
a proper addressing, raise concerns on reliability and fairness of the
intervention procedure.
| [
{
"version": "v1",
"created": "Tue, 28 Feb 2023 02:37:24 GMT"
},
{
"version": "v2",
"created": "Mon, 5 Jun 2023 01:40:49 GMT"
},
{
"version": "v3",
"created": "Mon, 3 Jul 2023 03:54:26 GMT"
}
] | 2023-07-04T00:00:00 | [
[
"Shin",
"Sungbin",
""
],
[
"Jo",
"Yohan",
""
],
[
"Ahn",
"Sungsoo",
""
],
[
"Lee",
"Namhoon",
""
]
] | TITLE: A Closer Look at the Intervention Procedure of Concept Bottleneck Models
ABSTRACT: Concept bottleneck models (CBMs) are a class of interpretable neural network
models that predict the target response of a given input based on its
high-level concepts. Unlike the standard end-to-end models, CBMs enable domain
experts to intervene on the predicted concepts and rectify any mistakes at test
time, so that more accurate task predictions can be made at the end. While such
intervenability provides a powerful avenue of control, many aspects of the
intervention procedure remain rather unexplored. In this work, we develop
various ways of selecting intervening concepts to improve the intervention
effectiveness and conduct an array of in-depth analyses as to how they evolve
under different circumstances. Specifically, we find that an informed
intervention strategy can reduce the task error more than ten times compared to
the current baseline under the same amount of intervention counts in realistic
settings, and yet, this can vary quite significantly when taking into account
different intervention granularity. We verify our findings through
comprehensive evaluations, not only on the standard real datasets, but also on
synthetic datasets that we generate based on a set of different causal graphs.
We further discover some major pitfalls of the current practices which, without
a proper addressing, raise concerns on reliability and fairness of the
intervention procedure.
| no_new_dataset | 0.707967 |
2303.02063 | Zhaobin Mo | Xuan Di, Rongye Shi, Zhaobin Mo, Yongjie Fu | Physics-Informed Deep Learning For Traffic State Estimation: A Survey
and the Outlook | null | null | 10.3390/a16060305 | null | cs.LG | http://creativecommons.org/licenses/by/4.0/ | For its robust predictive power (compared to pure physics-based models) and
sample-efficient training (compared to pure deep learning models),
physics-informed deep learning (PIDL), a paradigm hybridizing physics-based
models and deep neural networks (DNN), has been booming in science and
engineering fields. One key challenge of applying PIDL to various domains and
problems lies in the design of a computational graph that integrates physics
and DNNs. In other words, how physics are encoded into DNNs and how the physics
and data components are represented. In this paper, we provide a variety of
architecture designs of PIDL computational graphs and how these structures are
customized to traffic state estimation (TSE), a central problem in
transportation engineering. When observation data, problem type, and goal vary,
we demonstrate potential architectures of PIDL computational graphs and compare
these variants using the same real-world dataset.
| [
{
"version": "v1",
"created": "Fri, 3 Mar 2023 16:32:37 GMT"
},
{
"version": "v2",
"created": "Sat, 1 Jul 2023 18:59:06 GMT"
}
] | 2023-07-04T00:00:00 | [
[
"Di",
"Xuan",
""
],
[
"Shi",
"Rongye",
""
],
[
"Mo",
"Zhaobin",
""
],
[
"Fu",
"Yongjie",
""
]
] | TITLE: Physics-Informed Deep Learning For Traffic State Estimation: A Survey
and the Outlook
ABSTRACT: For its robust predictive power (compared to pure physics-based models) and
sample-efficient training (compared to pure deep learning models),
physics-informed deep learning (PIDL), a paradigm hybridizing physics-based
models and deep neural networks (DNN), has been booming in science and
engineering fields. One key challenge of applying PIDL to various domains and
problems lies in the design of a computational graph that integrates physics
and DNNs. In other words, how physics are encoded into DNNs and how the physics
and data components are represented. In this paper, we provide a variety of
architecture designs of PIDL computational graphs and how these structures are
customized to traffic state estimation (TSE), a central problem in
transportation engineering. When observation data, problem type, and goal vary,
we demonstrate potential architectures of PIDL computational graphs and compare
these variants using the same real-world dataset.
| no_new_dataset | 0.709403 |
2303.04253 | Lijing Zhu | Lijing Zhu, Qizhen Lan, Alvaro Velasquez, Houbing Song, Acharya Kamal,
Qing Tian, Shuteng Niu | TMHOI: Translational Model for Human-Object Interaction Detection | 10 pages, 3 figures, 2 tables | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Detecting human-object interactions (HOIs) is an intricate challenge in the
field of computer vision. Existing methods for HOI detection heavily rely on
appearance-based features, but these may not fully capture all the essential
characteristics necessary for accurate detection. To overcome these challenges,
we propose an innovative graph-based approach called TMGHOI (Translational
Model for Human-Object Interaction Detection). Our method effectively captures
the sentiment representation of HOIs by integrating both spatial and semantic
knowledge. By representing HOIs as a graph, where the interaction components
serve as nodes and their spatial relationships as edges. To extract crucial
spatial and semantic information, TMGHOI employs separate spatial and semantic
encoders. Subsequently, these encodings are combined to construct a knowledge
graph that effectively captures the sentiment representation of HOIs.
Additionally, the ability to incorporate prior knowledge enhances the
understanding of interactions, further boosting detection accuracy. We
conducted extensive evaluations on the widely-used HICO-DET datasets to
demonstrate the effectiveness of TMGHOI. Our approach outperformed existing
state-of-the-art graph-based methods by a significant margin, showcasing its
potential as a superior solution for HOI detection. We are confident that
TMGHOI has the potential to significantly improve the accuracy and efficiency
of HOI detection. Its integration of spatial and semantic knowledge, along with
its computational efficiency and practicality, makes it a valuable tool for
researchers and practitioners in the computer vision community. As with any
research, we acknowledge the importance of further exploration and evaluation
on various datasets to establish the generalizability and robustness of our
proposed method.
| [
{
"version": "v1",
"created": "Tue, 7 Mar 2023 21:52:10 GMT"
},
{
"version": "v2",
"created": "Wed, 15 Mar 2023 02:52:14 GMT"
},
{
"version": "v3",
"created": "Sat, 1 Jul 2023 15:44:42 GMT"
}
] | 2023-07-04T00:00:00 | [
[
"Zhu",
"Lijing",
""
],
[
"Lan",
"Qizhen",
""
],
[
"Velasquez",
"Alvaro",
""
],
[
"Song",
"Houbing",
""
],
[
"Kamal",
"Acharya",
""
],
[
"Tian",
"Qing",
""
],
[
"Niu",
"Shuteng",
""
]
] | TITLE: TMHOI: Translational Model for Human-Object Interaction Detection
ABSTRACT: Detecting human-object interactions (HOIs) is an intricate challenge in the
field of computer vision. Existing methods for HOI detection heavily rely on
appearance-based features, but these may not fully capture all the essential
characteristics necessary for accurate detection. To overcome these challenges,
we propose an innovative graph-based approach called TMGHOI (Translational
Model for Human-Object Interaction Detection). Our method effectively captures
the sentiment representation of HOIs by integrating both spatial and semantic
knowledge. By representing HOIs as a graph, where the interaction components
serve as nodes and their spatial relationships as edges. To extract crucial
spatial and semantic information, TMGHOI employs separate spatial and semantic
encoders. Subsequently, these encodings are combined to construct a knowledge
graph that effectively captures the sentiment representation of HOIs.
Additionally, the ability to incorporate prior knowledge enhances the
understanding of interactions, further boosting detection accuracy. We
conducted extensive evaluations on the widely-used HICO-DET datasets to
demonstrate the effectiveness of TMGHOI. Our approach outperformed existing
state-of-the-art graph-based methods by a significant margin, showcasing its
potential as a superior solution for HOI detection. We are confident that
TMGHOI has the potential to significantly improve the accuracy and efficiency
of HOI detection. Its integration of spatial and semantic knowledge, along with
its computational efficiency and practicality, makes it a valuable tool for
researchers and practitioners in the computer vision community. As with any
research, we acknowledge the importance of further exploration and evaluation
on various datasets to establish the generalizability and robustness of our
proposed method.
| no_new_dataset | 0.709837 |
2303.07399 | Yining Li | Tao Jiang, Peng Lu, Li Zhang, Ningsheng Ma, Rui Han, Chengqi Lyu,
Yining Li, Kai Chen | RTMPose: Real-Time Multi-Person Pose Estimation based on MMPose | null | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | Recent studies on 2D pose estimation have achieved excellent performance on
public benchmarks, yet its application in the industrial community still
suffers from heavy model parameters and high latency. In order to bridge this
gap, we empirically explore key factors in pose estimation including paradigm,
model architecture, training strategy, and deployment, and present a
high-performance real-time multi-person pose estimation framework, RTMPose,
based on MMPose. Our RTMPose-m achieves 75.8% AP on COCO with 90+ FPS on an
Intel i7-11700 CPU and 430+ FPS on an NVIDIA GTX 1660 Ti GPU, and RTMPose-l
achieves 67.0% AP on COCO-WholeBody with 130+ FPS. To further evaluate
RTMPose's capability in critical real-time applications, we also report the
performance after deploying on the mobile device. Our RTMPose-s achieves 72.2%
AP on COCO with 70+ FPS on a Snapdragon 865 chip, outperforming existing
open-source libraries. Code and models are released at
https://github.com/open-mmlab/mmpose/tree/1.x/projects/rtmpose.
| [
{
"version": "v1",
"created": "Mon, 13 Mar 2023 18:26:11 GMT"
},
{
"version": "v2",
"created": "Mon, 3 Jul 2023 03:06:26 GMT"
}
] | 2023-07-04T00:00:00 | [
[
"Jiang",
"Tao",
""
],
[
"Lu",
"Peng",
""
],
[
"Zhang",
"Li",
""
],
[
"Ma",
"Ningsheng",
""
],
[
"Han",
"Rui",
""
],
[
"Lyu",
"Chengqi",
""
],
[
"Li",
"Yining",
""
],
[
"Chen",
"Kai",
""
]
] | TITLE: RTMPose: Real-Time Multi-Person Pose Estimation based on MMPose
ABSTRACT: Recent studies on 2D pose estimation have achieved excellent performance on
public benchmarks, yet its application in the industrial community still
suffers from heavy model parameters and high latency. In order to bridge this
gap, we empirically explore key factors in pose estimation including paradigm,
model architecture, training strategy, and deployment, and present a
high-performance real-time multi-person pose estimation framework, RTMPose,
based on MMPose. Our RTMPose-m achieves 75.8% AP on COCO with 90+ FPS on an
Intel i7-11700 CPU and 430+ FPS on an NVIDIA GTX 1660 Ti GPU, and RTMPose-l
achieves 67.0% AP on COCO-WholeBody with 130+ FPS. To further evaluate
RTMPose's capability in critical real-time applications, we also report the
performance after deploying on the mobile device. Our RTMPose-s achieves 72.2%
AP on COCO with 70+ FPS on a Snapdragon 865 chip, outperforming existing
open-source libraries. Code and models are released at
https://github.com/open-mmlab/mmpose/tree/1.x/projects/rtmpose.
| no_new_dataset | 0.709384 |
2303.13915 | Theresa Bender | Theresa Bender, Philip Gemke, Ennio Idrobo-Avila, Henning Dathe,
Dagmar Krefting, Nicolai Spicher | Benchmarking the Impact of Noise on Deep Learning-based Classification
of Atrial Fibrillation in 12-Lead ECG | to be published in proceedings of Medical Informatics Europe 2023 | null | 10.3233/SHTI230321 | null | eess.SP cs.LG | http://creativecommons.org/licenses/by/4.0/ | Electrocardiography analysis is widely used in various clinical applications
and Deep Learning models for classification tasks are currently in the focus of
research. Due to their data-driven character, they bear the potential to handle
signal noise efficiently, but its influence on the accuracy of these methods is
still unclear. Therefore, we benchmark the influence of four types of noise on
the accuracy of a Deep Learning-based method for atrial fibrillation detection
in 12-lead electrocardiograms. We use a subset of a publicly available dataset
(PTBXL) and use the metadata provided by human experts regarding noise for
assigning a signal quality to each electrocardiogram. Furthermore, we compute a
quantitative signal-to-noise ratio for each electrocardiogram. We analyze the
accuracy of the Deep Learning model with respect to both metrics and observe
that the method can robustly identify atrial fibrillation, even in cases
signals are labelled by human experts as being noisy on multiple leads. False
positive and false negative rates are slightly worse for data being labelled as
noisy. Interestingly, data annotated as showing baseline drift noise results in
an accuracy very similar to data without. We conclude that the issue of
processing noisy electrocardiography data can be addressed successfully by Deep
Learning methods that might not need preprocessing as many conventional methods
do.
| [
{
"version": "v1",
"created": "Fri, 24 Mar 2023 11:04:16 GMT"
}
] | 2023-07-04T00:00:00 | [
[
"Bender",
"Theresa",
""
],
[
"Gemke",
"Philip",
""
],
[
"Idrobo-Avila",
"Ennio",
""
],
[
"Dathe",
"Henning",
""
],
[
"Krefting",
"Dagmar",
""
],
[
"Spicher",
"Nicolai",
""
]
] | TITLE: Benchmarking the Impact of Noise on Deep Learning-based Classification
of Atrial Fibrillation in 12-Lead ECG
ABSTRACT: Electrocardiography analysis is widely used in various clinical applications
and Deep Learning models for classification tasks are currently in the focus of
research. Due to their data-driven character, they bear the potential to handle
signal noise efficiently, but its influence on the accuracy of these methods is
still unclear. Therefore, we benchmark the influence of four types of noise on
the accuracy of a Deep Learning-based method for atrial fibrillation detection
in 12-lead electrocardiograms. We use a subset of a publicly available dataset
(PTBXL) and use the metadata provided by human experts regarding noise for
assigning a signal quality to each electrocardiogram. Furthermore, we compute a
quantitative signal-to-noise ratio for each electrocardiogram. We analyze the
accuracy of the Deep Learning model with respect to both metrics and observe
that the method can robustly identify atrial fibrillation, even in cases
signals are labelled by human experts as being noisy on multiple leads. False
positive and false negative rates are slightly worse for data being labelled as
noisy. Interestingly, data annotated as showing baseline drift noise results in
an accuracy very similar to data without. We conclude that the issue of
processing noisy electrocardiography data can be addressed successfully by Deep
Learning methods that might not need preprocessing as many conventional methods
do.
| no_new_dataset | 0.705405 |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.