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
|
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
2306.05107 | Quan Quan | Quan Quan, Shang Zhao, Qingsong Yao, Heqin Zhu, S. Kevin Zhou | Unsupervised augmentation optimization for few-shot medical image
segmentation | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The augmentation parameters matter to few-shot semantic segmentation since
they directly affect the training outcome by feeding the networks with varying
perturbated samples. However, searching optimal augmentation parameters for
few-shot segmentation models without annotations is a challenge that current
methods fail to address. In this paper, we first propose a framework to
determine the ``optimal'' parameters without human annotations by solving a
distribution-matching problem between the intra-instance and intra-class
similarity distribution, with the intra-instance similarity describing the
similarity between the original sample of a particular anatomy and its
augmented ones and the intra-class similarity representing the similarity
between the selected sample and the others in the same class. Extensive
experiments demonstrate the superiority of our optimized augmentation in
boosting few-shot segmentation models. We greatly improve the top competing
method by 1.27\% and 1.11\% on Abd-MRI and Abd-CT datasets, respectively, and
even achieve a significant improvement for SSL-ALP on the left kidney by 3.39\%
on the Abd-CT dataset.
| [
{
"version": "v1",
"created": "Thu, 8 Jun 2023 11:15:04 GMT"
}
] | 2023-06-09T00:00:00 | [
[
"Quan",
"Quan",
""
],
[
"Zhao",
"Shang",
""
],
[
"Yao",
"Qingsong",
""
],
[
"Zhu",
"Heqin",
""
],
[
"Zhou",
"S. Kevin",
""
]
] | TITLE: Unsupervised augmentation optimization for few-shot medical image
segmentation
ABSTRACT: The augmentation parameters matter to few-shot semantic segmentation since
they directly affect the training outcome by feeding the networks with varying
perturbated samples. However, searching optimal augmentation parameters for
few-shot segmentation models without annotations is a challenge that current
methods fail to address. In this paper, we first propose a framework to
determine the ``optimal'' parameters without human annotations by solving a
distribution-matching problem between the intra-instance and intra-class
similarity distribution, with the intra-instance similarity describing the
similarity between the original sample of a particular anatomy and its
augmented ones and the intra-class similarity representing the similarity
between the selected sample and the others in the same class. Extensive
experiments demonstrate the superiority of our optimized augmentation in
boosting few-shot segmentation models. We greatly improve the top competing
method by 1.27\% and 1.11\% on Abd-MRI and Abd-CT datasets, respectively, and
even achieve a significant improvement for SSL-ALP on the left kidney by 3.39\%
on the Abd-CT dataset.
|
2306.05116 | Christian Herold | Christian Herold and Hermann Ney | On Search Strategies for Document-Level Neural Machine Translation | Accepted to ACL 2023 (Findings) | null | null | null | cs.CL cs.AI cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Compared to sentence-level systems, document-level neural machine translation
(NMT) models produce a more consistent output across a document and are able to
better resolve ambiguities within the input. There are many works on
document-level NMT, mostly focusing on modifying the model architecture or
training strategy to better accommodate the additional context-input. On the
other hand, in most works, the question on how to perform search with the
trained model is scarcely discussed, sometimes not mentioned at all. In this
work, we aim to answer the question how to best utilize a context-aware
translation model in decoding. We start with the most popular document-level
NMT approach and compare different decoding schemes, some from the literature
and others proposed by us. In the comparison, we are using both, standard
automatic metrics, as well as specific linguistic phenomena on three standard
document-level translation benchmarks. We find that most commonly used decoding
strategies perform similar to each other and that higher quality context
information has the potential to further improve the translation.
| [
{
"version": "v1",
"created": "Thu, 8 Jun 2023 11:30:43 GMT"
}
] | 2023-06-09T00:00:00 | [
[
"Herold",
"Christian",
""
],
[
"Ney",
"Hermann",
""
]
] | TITLE: On Search Strategies for Document-Level Neural Machine Translation
ABSTRACT: Compared to sentence-level systems, document-level neural machine translation
(NMT) models produce a more consistent output across a document and are able to
better resolve ambiguities within the input. There are many works on
document-level NMT, mostly focusing on modifying the model architecture or
training strategy to better accommodate the additional context-input. On the
other hand, in most works, the question on how to perform search with the
trained model is scarcely discussed, sometimes not mentioned at all. In this
work, we aim to answer the question how to best utilize a context-aware
translation model in decoding. We start with the most popular document-level
NMT approach and compare different decoding schemes, some from the literature
and others proposed by us. In the comparison, we are using both, standard
automatic metrics, as well as specific linguistic phenomena on three standard
document-level translation benchmarks. We find that most commonly used decoding
strategies perform similar to each other and that higher quality context
information has the potential to further improve the translation.
|
2306.05119 | Mingqi Gao | Mingqi Gao, Xiaojun Wan, Jia Su, Zhefeng Wang, Baoxing Huai | Reference Matters: Benchmarking Factual Error Correction for Dialogue
Summarization with Fine-grained Evaluation Framework | Accepted to ACL 2023 Main Conference | null | null | null | cs.CL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Factuality is important to dialogue summarization. Factual error correction
(FEC) of model-generated summaries is one way to improve factuality. Current
FEC evaluation that relies on factuality metrics is not reliable and detailed
enough. To address this problem, we are the first to manually annotate a FEC
dataset for dialogue summarization containing 4000 items and propose FERRANTI,
a fine-grained evaluation framework based on reference correction that
automatically evaluates the performance of FEC models on different error
categories. Using this evaluation framework, we conduct sufficient experiments
with FEC approaches under a variety of settings and find the best training
modes and significant differences in the performance of the existing approaches
on different factual error categories.
| [
{
"version": "v1",
"created": "Thu, 8 Jun 2023 11:41:39 GMT"
}
] | 2023-06-09T00:00:00 | [
[
"Gao",
"Mingqi",
""
],
[
"Wan",
"Xiaojun",
""
],
[
"Su",
"Jia",
""
],
[
"Wang",
"Zhefeng",
""
],
[
"Huai",
"Baoxing",
""
]
] | TITLE: Reference Matters: Benchmarking Factual Error Correction for Dialogue
Summarization with Fine-grained Evaluation Framework
ABSTRACT: Factuality is important to dialogue summarization. Factual error correction
(FEC) of model-generated summaries is one way to improve factuality. Current
FEC evaluation that relies on factuality metrics is not reliable and detailed
enough. To address this problem, we are the first to manually annotate a FEC
dataset for dialogue summarization containing 4000 items and propose FERRANTI,
a fine-grained evaluation framework based on reference correction that
automatically evaluates the performance of FEC models on different error
categories. Using this evaluation framework, we conduct sufficient experiments
with FEC approaches under a variety of settings and find the best training
modes and significant differences in the performance of the existing approaches
on different factual error categories.
|
2306.05122 | Henrik Bj{\o}rn Axelsen | Henrik Axelsen, Johannes Rude Jensen, Sebastian Axelsen, Valdemar
Licht, Omri Ross | Can AI Moderate Online Communities? | null | null | null | null | cs.CY | http://creativecommons.org/licenses/by/4.0/ | The task of cultivating healthy communication in online communities becomes
increasingly urgent, as gaming and social media experiences become
progressively more immersive and life-like. We approach the challenge of
moderating online communities by training student models using a large language
model (LLM). We use zero-shot learning models to distill and expand datasets
followed by a few-shot learning and a fine-tuning approach, leveraging
open-access generative pre-trained transformer models (GPT) from OpenAI. Our
preliminary findings suggest, that when properly trained, LLMs can excel in
identifying actor intentions, moderating toxic comments, and rewarding positive
contributions. The student models perform above-expectation in non-contextual
assignments such as identifying classically toxic behavior and perform
sufficiently on contextual assignments such as identifying positive
contributions to online discourse. Further, using open-access models like
OpenAI's GPT we experience a step-change in the development process for what
has historically been a complex modeling task. We contribute to the information
system (IS) discourse with a rapid development framework on the application of
generative AI in content online moderation and management of culture in
decentralized, pseudonymous communities by providing a sample model suite of
industrial-ready generative AI models based on open-access LLMs.
| [
{
"version": "v1",
"created": "Thu, 8 Jun 2023 11:45:44 GMT"
}
] | 2023-06-09T00:00:00 | [
[
"Axelsen",
"Henrik",
""
],
[
"Jensen",
"Johannes Rude",
""
],
[
"Axelsen",
"Sebastian",
""
],
[
"Licht",
"Valdemar",
""
],
[
"Ross",
"Omri",
""
]
] | TITLE: Can AI Moderate Online Communities?
ABSTRACT: The task of cultivating healthy communication in online communities becomes
increasingly urgent, as gaming and social media experiences become
progressively more immersive and life-like. We approach the challenge of
moderating online communities by training student models using a large language
model (LLM). We use zero-shot learning models to distill and expand datasets
followed by a few-shot learning and a fine-tuning approach, leveraging
open-access generative pre-trained transformer models (GPT) from OpenAI. Our
preliminary findings suggest, that when properly trained, LLMs can excel in
identifying actor intentions, moderating toxic comments, and rewarding positive
contributions. The student models perform above-expectation in non-contextual
assignments such as identifying classically toxic behavior and perform
sufficiently on contextual assignments such as identifying positive
contributions to online discourse. Further, using open-access models like
OpenAI's GPT we experience a step-change in the development process for what
has historically been a complex modeling task. We contribute to the information
system (IS) discourse with a rapid development framework on the application of
generative AI in content online moderation and management of culture in
decentralized, pseudonymous communities by providing a sample model suite of
industrial-ready generative AI models based on open-access LLMs.
|
2306.05123 | Fouad Oubari | Fouad Oubari, Raphael Meunier, Rodrigue D\'ecatoire, Mathilde Mougeot | A Meta-Generation framework for Industrial System Generation | null | null | null | null | cs.LG | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Generative design is an increasingly important tool in the industrial world.
It allows the designers and engineers to easily explore vast ranges of design
options, providing a cheaper and faster alternative to the trial and failure
approaches. Thanks to the flexibility they offer, Deep Generative Models are
gaining popularity amongst Generative Design technologies. However, developing
and evaluating these models can be challenging. The field lacks accessible
benchmarks, in order to evaluate and compare objectively different Deep
Generative Models architectures. Moreover, vanilla Deep Generative Models
appear to be unable to accurately generate multi-components industrial systems
that are controlled by latent design constraints. To address these challenges,
we propose an industry-inspired use case that incorporates actual industrial
system characteristics. This use case can be quickly generated and used as a
benchmark. We propose a Meta-VAE capable of producing multi-component
industrial systems and showcase its application on the proposed use case.
| [
{
"version": "v1",
"created": "Thu, 8 Jun 2023 11:47:02 GMT"
}
] | 2023-06-09T00:00:00 | [
[
"Oubari",
"Fouad",
""
],
[
"Meunier",
"Raphael",
""
],
[
"Décatoire",
"Rodrigue",
""
],
[
"Mougeot",
"Mathilde",
""
]
] | TITLE: A Meta-Generation framework for Industrial System Generation
ABSTRACT: Generative design is an increasingly important tool in the industrial world.
It allows the designers and engineers to easily explore vast ranges of design
options, providing a cheaper and faster alternative to the trial and failure
approaches. Thanks to the flexibility they offer, Deep Generative Models are
gaining popularity amongst Generative Design technologies. However, developing
and evaluating these models can be challenging. The field lacks accessible
benchmarks, in order to evaluate and compare objectively different Deep
Generative Models architectures. Moreover, vanilla Deep Generative Models
appear to be unable to accurately generate multi-components industrial systems
that are controlled by latent design constraints. To address these challenges,
we propose an industry-inspired use case that incorporates actual industrial
system characteristics. This use case can be quickly generated and used as a
benchmark. We propose a Meta-VAE capable of producing multi-component
industrial systems and showcase its application on the proposed use case.
|
2306.05126 | Antonia Karamolegkou | Antonia Karamolegkou, Mostafa Abdou, Anders S{\o}gaard | Mapping Brains with Language Models: A Survey | null | null | null | null | cs.CL | http://creativecommons.org/licenses/by/4.0/ | Over the years, many researchers have seemingly made the same observation:
Brain and language model activations exhibit some structural similarities,
enabling linear partial mappings between features extracted from neural
recordings and computational language models. In an attempt to evaluate how
much evidence has been accumulated for this observation, we survey over 30
studies spanning 10 datasets and 8 metrics. How much evidence has been
accumulated, and what, if anything, is missing before we can draw conclusions?
Our analysis of the evaluation methods used in the literature reveals that some
of the metrics are less conservative. We also find that the accumulated
evidence, for now, remains ambiguous, but correlations with model size and
quality provide grounds for cautious optimism.
| [
{
"version": "v1",
"created": "Thu, 8 Jun 2023 11:50:58 GMT"
}
] | 2023-06-09T00:00:00 | [
[
"Karamolegkou",
"Antonia",
""
],
[
"Abdou",
"Mostafa",
""
],
[
"Søgaard",
"Anders",
""
]
] | TITLE: Mapping Brains with Language Models: A Survey
ABSTRACT: Over the years, many researchers have seemingly made the same observation:
Brain and language model activations exhibit some structural similarities,
enabling linear partial mappings between features extracted from neural
recordings and computational language models. In an attempt to evaluate how
much evidence has been accumulated for this observation, we survey over 30
studies spanning 10 datasets and 8 metrics. How much evidence has been
accumulated, and what, if anything, is missing before we can draw conclusions?
Our analysis of the evaluation methods used in the literature reveals that some
of the metrics are less conservative. We also find that the accumulated
evidence, for now, remains ambiguous, but correlations with model size and
quality provide grounds for cautious optimism.
|
2306.05129 | Zenglin Shi | Zenglin Shi, Pascal Mettes, Cees G.M. Snoek | Focus for Free in Density-Based Counting | 18 pages | null | null | null | cs.CV | http://creativecommons.org/licenses/by-nc-nd/4.0/ | This work considers supervised learning to count from images and their
corresponding point annotations. Where density-based counting methods typically
use the point annotations only to create Gaussian-density maps, which act as
the supervision signal, the starting point of this work is that point
annotations have counting potential beyond density map generation. We introduce
two methods that repurpose the available point annotations to enhance counting
performance. The first is a counting-specific augmentation that leverages point
annotations to simulate occluded objects in both input and density images to
enhance the network's robustness to occlusions. The second method, foreground
distillation, generates foreground masks from the point annotations, from which
we train an auxiliary network on images with blacked-out backgrounds. By doing
so, it learns to extract foreground counting knowledge without interference
from the background. These methods can be seamlessly integrated with existing
counting advances and are adaptable to different loss functions. We demonstrate
complementary effects of the approaches, allowing us to achieve robust counting
results even in challenging scenarios such as background clutter, occlusion,
and varying crowd densities. Our proposed approach achieves strong counting
results on multiple datasets, including ShanghaiTech Part\_A and Part\_B,
UCF\_QNRF, JHU-Crowd++, and NWPU-Crowd.
| [
{
"version": "v1",
"created": "Thu, 8 Jun 2023 11:54:37 GMT"
}
] | 2023-06-09T00:00:00 | [
[
"Shi",
"Zenglin",
""
],
[
"Mettes",
"Pascal",
""
],
[
"Snoek",
"Cees G. M.",
""
]
] | TITLE: Focus for Free in Density-Based Counting
ABSTRACT: This work considers supervised learning to count from images and their
corresponding point annotations. Where density-based counting methods typically
use the point annotations only to create Gaussian-density maps, which act as
the supervision signal, the starting point of this work is that point
annotations have counting potential beyond density map generation. We introduce
two methods that repurpose the available point annotations to enhance counting
performance. The first is a counting-specific augmentation that leverages point
annotations to simulate occluded objects in both input and density images to
enhance the network's robustness to occlusions. The second method, foreground
distillation, generates foreground masks from the point annotations, from which
we train an auxiliary network on images with blacked-out backgrounds. By doing
so, it learns to extract foreground counting knowledge without interference
from the background. These methods can be seamlessly integrated with existing
counting advances and are adaptable to different loss functions. We demonstrate
complementary effects of the approaches, allowing us to achieve robust counting
results even in challenging scenarios such as background clutter, occlusion,
and varying crowd densities. Our proposed approach achieves strong counting
results on multiple datasets, including ShanghaiTech Part\_A and Part\_B,
UCF\_QNRF, JHU-Crowd++, and NWPU-Crowd.
|
2306.05135 | H{\aa}kon Hukkel{\aa}s | H{\aa}kon Hukkel{\aa}s, Frank Lindseth | Does Image Anonymization Impact Computer Vision Training? | Accepted at CVPR Workshop on Autonomous Driving 2023 | null | null | null | cs.CV cs.AI | http://creativecommons.org/licenses/by-sa/4.0/ | Image anonymization is widely adapted in practice to comply with privacy
regulations in many regions. However, anonymization often degrades the quality
of the data, reducing its utility for computer vision development. In this
paper, we investigate the impact of image anonymization for training computer
vision models on key computer vision tasks (detection, instance segmentation,
and pose estimation). Specifically, we benchmark the recognition drop on common
detection datasets, where we evaluate both traditional and realistic
anonymization for faces and full bodies. Our comprehensive experiments reflect
that traditional image anonymization substantially impacts final model
performance, particularly when anonymizing the full body. Furthermore, we find
that realistic anonymization can mitigate this decrease in performance, where
our experiments reflect a minimal performance drop for face anonymization. Our
study demonstrates that realistic anonymization can enable privacy-preserving
computer vision development with minimal performance degradation across a range
of important computer vision benchmarks.
| [
{
"version": "v1",
"created": "Thu, 8 Jun 2023 12:02:03 GMT"
}
] | 2023-06-09T00:00:00 | [
[
"Hukkelås",
"Håkon",
""
],
[
"Lindseth",
"Frank",
""
]
] | TITLE: Does Image Anonymization Impact Computer Vision Training?
ABSTRACT: Image anonymization is widely adapted in practice to comply with privacy
regulations in many regions. However, anonymization often degrades the quality
of the data, reducing its utility for computer vision development. In this
paper, we investigate the impact of image anonymization for training computer
vision models on key computer vision tasks (detection, instance segmentation,
and pose estimation). Specifically, we benchmark the recognition drop on common
detection datasets, where we evaluate both traditional and realistic
anonymization for faces and full bodies. Our comprehensive experiments reflect
that traditional image anonymization substantially impacts final model
performance, particularly when anonymizing the full body. Furthermore, we find
that realistic anonymization can mitigate this decrease in performance, where
our experiments reflect a minimal performance drop for face anonymization. Our
study demonstrates that realistic anonymization can enable privacy-preserving
computer vision development with minimal performance degradation across a range
of important computer vision benchmarks.
|
2306.05196 | Hejun Huang | Hejun Huang, Zuguo Chen, Ying Zou, Ming Lu, Chaoyang Chen | Channel prior convolutional attention for medical image segmentation | null | null | null | null | eess.IV cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Characteristics such as low contrast and significant organ shape variations
are often exhibited in medical images. The improvement of segmentation
performance in medical imaging is limited by the generally insufficient
adaptive capabilities of existing attention mechanisms. An efficient Channel
Prior Convolutional Attention (CPCA) method is proposed in this paper,
supporting the dynamic distribution of attention weights in both channel and
spatial dimensions. Spatial relationships are effectively extracted while
preserving the channel prior by employing a multi-scale depth-wise
convolutional module. The ability to focus on informative channels and
important regions is possessed by CPCA. A segmentation network called CPCANet
for medical image segmentation is proposed based on CPCA. CPCANet is validated
on two publicly available datasets. Improved segmentation performance is
achieved by CPCANet while requiring fewer computational resources through
comparisons with state-of-the-art algorithms. Our code is publicly available at
\url{https://github.com/Cuthbert-Huang/CPCANet}.
| [
{
"version": "v1",
"created": "Thu, 8 Jun 2023 13:52:41 GMT"
}
] | 2023-06-09T00:00:00 | [
[
"Huang",
"Hejun",
""
],
[
"Chen",
"Zuguo",
""
],
[
"Zou",
"Ying",
""
],
[
"Lu",
"Ming",
""
],
[
"Chen",
"Chaoyang",
""
]
] | TITLE: Channel prior convolutional attention for medical image segmentation
ABSTRACT: Characteristics such as low contrast and significant organ shape variations
are often exhibited in medical images. The improvement of segmentation
performance in medical imaging is limited by the generally insufficient
adaptive capabilities of existing attention mechanisms. An efficient Channel
Prior Convolutional Attention (CPCA) method is proposed in this paper,
supporting the dynamic distribution of attention weights in both channel and
spatial dimensions. Spatial relationships are effectively extracted while
preserving the channel prior by employing a multi-scale depth-wise
convolutional module. The ability to focus on informative channels and
important regions is possessed by CPCA. A segmentation network called CPCANet
for medical image segmentation is proposed based on CPCA. CPCANet is validated
on two publicly available datasets. Improved segmentation performance is
achieved by CPCANet while requiring fewer computational resources through
comparisons with state-of-the-art algorithms. Our code is publicly available at
\url{https://github.com/Cuthbert-Huang/CPCANet}.
|
2306.05233 | Hailong Hu | Hailong Hu, Jun Pang | Ownership Protection of Generative Adversarial Networks | null | null | null | null | cs.CR cs.CV cs.LG | http://creativecommons.org/licenses/by/4.0/ | Generative adversarial networks (GANs) have shown remarkable success in image
synthesis, making GAN models themselves commercially valuable to legitimate
model owners. Therefore, it is critical to technically protect the intellectual
property of GANs. Prior works need to tamper with the training set or training
process, and they are not robust to emerging model extraction attacks. In this
paper, we propose a new ownership protection method based on the common
characteristics of a target model and its stolen models. Our method can be
directly applicable to all well-trained GANs as it does not require retraining
target models. Extensive experimental results show that our new method can
achieve the best protection performance, compared to the state-of-the-art
methods. Finally, we demonstrate the effectiveness of our method with respect
to the number of generations of model extraction attacks, the number of
generated samples, different datasets, as well as adaptive attacks.
| [
{
"version": "v1",
"created": "Thu, 8 Jun 2023 14:31:58 GMT"
}
] | 2023-06-09T00:00:00 | [
[
"Hu",
"Hailong",
""
],
[
"Pang",
"Jun",
""
]
] | TITLE: Ownership Protection of Generative Adversarial Networks
ABSTRACT: Generative adversarial networks (GANs) have shown remarkable success in image
synthesis, making GAN models themselves commercially valuable to legitimate
model owners. Therefore, it is critical to technically protect the intellectual
property of GANs. Prior works need to tamper with the training set or training
process, and they are not robust to emerging model extraction attacks. In this
paper, we propose a new ownership protection method based on the common
characteristics of a target model and its stolen models. Our method can be
directly applicable to all well-trained GANs as it does not require retraining
target models. Extensive experimental results show that our new method can
achieve the best protection performance, compared to the state-of-the-art
methods. Finally, we demonstrate the effectiveness of our method with respect
to the number of generations of model extraction attacks, the number of
generated samples, different datasets, as well as adaptive attacks.
|
2306.05241 | Peng Qi | Peng Qi, Yuyang Zhao, Yufeng Shen, Wei Ji, Juan Cao, Tat-Seng Chua | Two Heads Are Better Than One: Improving Fake News Video Detection by
Correlating with Neighbors | To appear in ACL 2023 Findings | null | null | null | cs.MM | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The prevalence of short video platforms has spawned a lot of fake news
videos, which have stronger propagation ability than textual fake news. Thus,
automatically detecting fake news videos has been an important countermeasure
in practice. Previous works commonly verify each news video individually with
multimodal information. Nevertheless, news videos from different perspectives
regarding the same event are commonly posted together, which contain
complementary or contradictory information and thus can be used to evaluate
each other mutually. To this end, we introduce a new and practical paradigm,
i.e., cross-sample fake news video detection, and propose a novel framework,
Neighbor-Enhanced fakE news video Detection (NEED), which integrates the
neighborhood relationship of new videos belonging to the same event. NEED can
be readily combined with existing single-sample detectors and further enhance
their performances with the proposed graph aggregation (GA) and debunking
rectification (DR) modules. Specifically, given the feature representations
obtained from single-sample detectors, GA aggregates the neighborhood
information with the dynamic graph to enrich the features of independent
samples. After that, DR explicitly leverages the relationship between debunking
videos and fake news videos to refute the candidate videos via textual and
visual consistency. Extensive experiments on the public benchmark demonstrate
that NEED greatly improves the performance of both single-modal (up to 8.34% in
accuracy) and multimodal (up to 4.97% in accuracy) base detectors. Codes are
available in https://github.com/ICTMCG/NEED.
| [
{
"version": "v1",
"created": "Thu, 8 Jun 2023 14:40:57 GMT"
}
] | 2023-06-09T00:00:00 | [
[
"Qi",
"Peng",
""
],
[
"Zhao",
"Yuyang",
""
],
[
"Shen",
"Yufeng",
""
],
[
"Ji",
"Wei",
""
],
[
"Cao",
"Juan",
""
],
[
"Chua",
"Tat-Seng",
""
]
] | TITLE: Two Heads Are Better Than One: Improving Fake News Video Detection by
Correlating with Neighbors
ABSTRACT: The prevalence of short video platforms has spawned a lot of fake news
videos, which have stronger propagation ability than textual fake news. Thus,
automatically detecting fake news videos has been an important countermeasure
in practice. Previous works commonly verify each news video individually with
multimodal information. Nevertheless, news videos from different perspectives
regarding the same event are commonly posted together, which contain
complementary or contradictory information and thus can be used to evaluate
each other mutually. To this end, we introduce a new and practical paradigm,
i.e., cross-sample fake news video detection, and propose a novel framework,
Neighbor-Enhanced fakE news video Detection (NEED), which integrates the
neighborhood relationship of new videos belonging to the same event. NEED can
be readily combined with existing single-sample detectors and further enhance
their performances with the proposed graph aggregation (GA) and debunking
rectification (DR) modules. Specifically, given the feature representations
obtained from single-sample detectors, GA aggregates the neighborhood
information with the dynamic graph to enrich the features of independent
samples. After that, DR explicitly leverages the relationship between debunking
videos and fake news videos to refute the candidate videos via textual and
visual consistency. Extensive experiments on the public benchmark demonstrate
that NEED greatly improves the performance of both single-modal (up to 8.34% in
accuracy) and multimodal (up to 4.97% in accuracy) base detectors. Codes are
available in https://github.com/ICTMCG/NEED.
|
2306.05242 | S\"ohnke Benedikt Fischedick | S\"ohnke Benedikt Fischedick, Daniel Seichter, Robin Schmidt, Leonard
Rabes, and Horst-Michael Gross | Efficient Multi-Task Scene Analysis with RGB-D Transformers | To be published in IEEE International Joint Conference on Neural
Networks (IJCNN) 2023 | null | null | null | cs.CV cs.RO | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Scene analysis is essential for enabling autonomous systems, such as mobile
robots, to operate in real-world environments. However, obtaining a
comprehensive understanding of the scene requires solving multiple tasks, such
as panoptic segmentation, instance orientation estimation, and scene
classification. Solving these tasks given limited computing and battery
capabilities on mobile platforms is challenging. To address this challenge, we
introduce an efficient multi-task scene analysis approach, called EMSAFormer,
that uses an RGB-D Transformer-based encoder to simultaneously perform the
aforementioned tasks. Our approach builds upon the previously published
EMSANet. However, we show that the dual CNN-based encoder of EMSANet can be
replaced with a single Transformer-based encoder. To achieve this, we
investigate how information from both RGB and depth data can be effectively
incorporated in a single encoder. To accelerate inference on robotic hardware,
we provide a custom NVIDIA TensorRT extension enabling highly optimization for
our EMSAFormer approach. Through extensive experiments on the commonly used
indoor datasets NYUv2, SUNRGB-D, and ScanNet, we show that our approach
achieves state-of-the-art performance while still enabling inference with up to
39.1 FPS on an NVIDIA Jetson AGX Orin 32 GB.
| [
{
"version": "v1",
"created": "Thu, 8 Jun 2023 14:41:56 GMT"
}
] | 2023-06-09T00:00:00 | [
[
"Fischedick",
"Söhnke Benedikt",
""
],
[
"Seichter",
"Daniel",
""
],
[
"Schmidt",
"Robin",
""
],
[
"Rabes",
"Leonard",
""
],
[
"Gross",
"Horst-Michael",
""
]
] | TITLE: Efficient Multi-Task Scene Analysis with RGB-D Transformers
ABSTRACT: Scene analysis is essential for enabling autonomous systems, such as mobile
robots, to operate in real-world environments. However, obtaining a
comprehensive understanding of the scene requires solving multiple tasks, such
as panoptic segmentation, instance orientation estimation, and scene
classification. Solving these tasks given limited computing and battery
capabilities on mobile platforms is challenging. To address this challenge, we
introduce an efficient multi-task scene analysis approach, called EMSAFormer,
that uses an RGB-D Transformer-based encoder to simultaneously perform the
aforementioned tasks. Our approach builds upon the previously published
EMSANet. However, we show that the dual CNN-based encoder of EMSANet can be
replaced with a single Transformer-based encoder. To achieve this, we
investigate how information from both RGB and depth data can be effectively
incorporated in a single encoder. To accelerate inference on robotic hardware,
we provide a custom NVIDIA TensorRT extension enabling highly optimization for
our EMSAFormer approach. Through extensive experiments on the commonly used
indoor datasets NYUv2, SUNRGB-D, and ScanNet, we show that our approach
achieves state-of-the-art performance while still enabling inference with up to
39.1 FPS on an NVIDIA Jetson AGX Orin 32 GB.
|
2306.05245 | Kumari Nishu | Kumari Nishu, Minsik Cho, Devang Naik | Matching Latent Encoding for Audio-Text based Keyword Spotting | null | null | null | null | eess.AS cs.LG cs.SD | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Using audio and text embeddings jointly for Keyword Spotting (KWS) has shown
high-quality results, but the key challenge of how to semantically align two
embeddings for multi-word keywords of different sequence lengths remains
largely unsolved. In this paper, we propose an audio-text-based end-to-end
model architecture for flexible keyword spotting (KWS), which builds upon
learned audio and text embeddings. Our architecture uses a novel dynamic
programming-based algorithm, Dynamic Sequence Partitioning (DSP), to optimally
partition the audio sequence into the same length as the word-based text
sequence using the monotonic alignment of spoken content. Our proposed model
consists of an encoder block to get audio and text embeddings, a projector
block to project individual embeddings to a common latent space, and an
audio-text aligner containing a novel DSP algorithm, which aligns the audio and
text embeddings to determine if the spoken content is the same as the text.
Experimental results show that our DSP is more effective than other
partitioning schemes, and the proposed architecture outperformed the
state-of-the-art results on the public dataset in terms of Area Under the ROC
Curve (AUC) and Equal-Error-Rate (EER) by 14.4 % and 28.9%, respectively.
| [
{
"version": "v1",
"created": "Thu, 8 Jun 2023 14:44:23 GMT"
}
] | 2023-06-09T00:00:00 | [
[
"Nishu",
"Kumari",
""
],
[
"Cho",
"Minsik",
""
],
[
"Naik",
"Devang",
""
]
] | TITLE: Matching Latent Encoding for Audio-Text based Keyword Spotting
ABSTRACT: Using audio and text embeddings jointly for Keyword Spotting (KWS) has shown
high-quality results, but the key challenge of how to semantically align two
embeddings for multi-word keywords of different sequence lengths remains
largely unsolved. In this paper, we propose an audio-text-based end-to-end
model architecture for flexible keyword spotting (KWS), which builds upon
learned audio and text embeddings. Our architecture uses a novel dynamic
programming-based algorithm, Dynamic Sequence Partitioning (DSP), to optimally
partition the audio sequence into the same length as the word-based text
sequence using the monotonic alignment of spoken content. Our proposed model
consists of an encoder block to get audio and text embeddings, a projector
block to project individual embeddings to a common latent space, and an
audio-text aligner containing a novel DSP algorithm, which aligns the audio and
text embeddings to determine if the spoken content is the same as the text.
Experimental results show that our DSP is more effective than other
partitioning schemes, and the proposed architecture outperformed the
state-of-the-art results on the public dataset in terms of Area Under the ROC
Curve (AUC) and Equal-Error-Rate (EER) by 14.4 % and 28.9%, respectively.
|
2306.05270 | Yanjun Gao | Yanjun Gao, Dmitriy Dligach, Timothy Miller, Matthew M. Churpek, Majid
Afshar | Overview of the Problem List Summarization (ProbSum) 2023 Shared Task on
Summarizing Patients' Active Diagnoses and Problems from Electronic Health
Record Progress Notes | To appear in the Proceedings of the 5th BioNLP Workshop at ACL | null | null | null | cs.CL | http://creativecommons.org/licenses/by-nc-sa/4.0/ | The BioNLP Workshop 2023 initiated the launch of a shared task on Problem
List Summarization (ProbSum) in January 2023. The aim of this shared task is to
attract future research efforts in building NLP models for real-world
diagnostic decision support applications, where a system generating relevant
and accurate diagnoses will augment the healthcare providers decision-making
process and improve the quality of care for patients. The goal for participants
is to develop models that generated a list of diagnoses and problems using
input from the daily care notes collected from the hospitalization of
critically ill patients. Eight teams submitted their final systems to the
shared task leaderboard. In this paper, we describe the tasks, datasets,
evaluation metrics, and baseline systems. Additionally, the techniques and
results of the evaluation of the different approaches tried by the
participating teams are summarized.
| [
{
"version": "v1",
"created": "Thu, 8 Jun 2023 15:19:57 GMT"
}
] | 2023-06-09T00:00:00 | [
[
"Gao",
"Yanjun",
""
],
[
"Dligach",
"Dmitriy",
""
],
[
"Miller",
"Timothy",
""
],
[
"Churpek",
"Matthew M.",
""
],
[
"Afshar",
"Majid",
""
]
] | TITLE: Overview of the Problem List Summarization (ProbSum) 2023 Shared Task on
Summarizing Patients' Active Diagnoses and Problems from Electronic Health
Record Progress Notes
ABSTRACT: The BioNLP Workshop 2023 initiated the launch of a shared task on Problem
List Summarization (ProbSum) in January 2023. The aim of this shared task is to
attract future research efforts in building NLP models for real-world
diagnostic decision support applications, where a system generating relevant
and accurate diagnoses will augment the healthcare providers decision-making
process and improve the quality of care for patients. The goal for participants
is to develop models that generated a list of diagnoses and problems using
input from the daily care notes collected from the hospitalization of
critically ill patients. Eight teams submitted their final systems to the
shared task leaderboard. In this paper, we describe the tasks, datasets,
evaluation metrics, and baseline systems. Additionally, the techniques and
results of the evaluation of the different approaches tried by the
participating teams are summarized.
|
2306.05276 | Enrico Santus | Simone Scaboro, Beatrice Portellia, Emmanuele Chersoni, Enrico Santus,
Giuseppe Serra | Extensive Evaluation of Transformer-based Architectures for Adverse Drug
Events Extraction | null | null | null | null | cs.CL cs.AI | http://creativecommons.org/licenses/by/4.0/ | Adverse Event (ADE) extraction is one of the core tasks in digital
pharmacovigilance, especially when applied to informal texts. This task has
been addressed by the Natural Language Processing community using large
pre-trained language models, such as BERT. Despite the great number of
Transformer-based architectures used in the literature, it is unclear which of
them has better performances and why. Therefore, in this paper we perform an
extensive evaluation and analysis of 19 Transformer-based models for ADE
extraction on informal texts. We compare the performance of all the considered
models on two datasets with increasing levels of informality (forums posts and
tweets). We also combine the purely Transformer-based models with two
commonly-used additional processing layers (CRF and LSTM), and analyze their
effect on the models performance. Furthermore, we use a well-established
feature importance technique (SHAP) to correlate the performance of the models
with a set of features that describe them: model category (AutoEncoding,
AutoRegressive, Text-to-Text), pretraining domain, training from scratch, and
model size in number of parameters. At the end of our analyses, we identify a
list of take-home messages that can be derived from the experimental data.
| [
{
"version": "v1",
"created": "Thu, 8 Jun 2023 15:25:24 GMT"
}
] | 2023-06-09T00:00:00 | [
[
"Scaboro",
"Simone",
""
],
[
"Portellia",
"Beatrice",
""
],
[
"Chersoni",
"Emmanuele",
""
],
[
"Santus",
"Enrico",
""
],
[
"Serra",
"Giuseppe",
""
]
] | TITLE: Extensive Evaluation of Transformer-based Architectures for Adverse Drug
Events Extraction
ABSTRACT: Adverse Event (ADE) extraction is one of the core tasks in digital
pharmacovigilance, especially when applied to informal texts. This task has
been addressed by the Natural Language Processing community using large
pre-trained language models, such as BERT. Despite the great number of
Transformer-based architectures used in the literature, it is unclear which of
them has better performances and why. Therefore, in this paper we perform an
extensive evaluation and analysis of 19 Transformer-based models for ADE
extraction on informal texts. We compare the performance of all the considered
models on two datasets with increasing levels of informality (forums posts and
tweets). We also combine the purely Transformer-based models with two
commonly-used additional processing layers (CRF and LSTM), and analyze their
effect on the models performance. Furthermore, we use a well-established
feature importance technique (SHAP) to correlate the performance of the models
with a set of features that describe them: model category (AutoEncoding,
AutoRegressive, Text-to-Text), pretraining domain, training from scratch, and
model size in number of parameters. At the end of our analyses, we identify a
list of take-home messages that can be derived from the experimental data.
|
2306.05279 | Zhiyun Fan | Zhiyun Fan, Linhao Dong, Chen Shen, Zhenlin Liang, Jun Zhang, Lu Lu,
Zejun Ma | Language-specific Acoustic Boundary Learning for Mandarin-English
Code-switching Speech Recognition | null | null | null | null | cs.SD | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Code-switching speech recognition (CSSR) transcribes speech that switches
between multiple languages or dialects within a single sentence. The main
challenge in this task is that different languages often have similar
pronunciations, making it difficult for models to distinguish between them. In
this paper, we propose a method for solving the CSSR task from the perspective
of language-specific acoustic boundary learning. We introduce language-specific
weight estimators (LSWE) to model acoustic boundary learning in different
languages separately. Additionally, a non-autoregressive (NAR) decoder and a
language change detection (LCD) module are employed to assist in training.
Evaluated on the SEAME corpus, our method achieves a state-of-the-art mixed
error rate (MER) of 16.29% and 22.81% on the test_man and test_sge sets. We
also demonstrate the effectiveness of our method on a 9000-hour in-house
meeting code-switching dataset, where our method achieves a relatively 7.9% MER
reduction.
| [
{
"version": "v1",
"created": "Thu, 8 Jun 2023 15:27:40 GMT"
}
] | 2023-06-09T00:00:00 | [
[
"Fan",
"Zhiyun",
""
],
[
"Dong",
"Linhao",
""
],
[
"Shen",
"Chen",
""
],
[
"Liang",
"Zhenlin",
""
],
[
"Zhang",
"Jun",
""
],
[
"Lu",
"Lu",
""
],
[
"Ma",
"Zejun",
""
]
] | TITLE: Language-specific Acoustic Boundary Learning for Mandarin-English
Code-switching Speech Recognition
ABSTRACT: Code-switching speech recognition (CSSR) transcribes speech that switches
between multiple languages or dialects within a single sentence. The main
challenge in this task is that different languages often have similar
pronunciations, making it difficult for models to distinguish between them. In
this paper, we propose a method for solving the CSSR task from the perspective
of language-specific acoustic boundary learning. We introduce language-specific
weight estimators (LSWE) to model acoustic boundary learning in different
languages separately. Additionally, a non-autoregressive (NAR) decoder and a
language change detection (LCD) module are employed to assist in training.
Evaluated on the SEAME corpus, our method achieves a state-of-the-art mixed
error rate (MER) of 16.29% and 22.81% on the test_man and test_sge sets. We
also demonstrate the effectiveness of our method on a 9000-hour in-house
meeting code-switching dataset, where our method achieves a relatively 7.9% MER
reduction.
|
2306.05283 | Chihoon Lee | Dixon Vimalajeewa, Chihoon Lee, Brani Vidakovic | A Method for Detecting Murmurous Heart Sounds based on Self-similar
Properties | null | null | null | null | eess.SP cs.LG | http://creativecommons.org/licenses/by/4.0/ | A heart murmur is an atypical sound produced by the flow of blood through the
heart. It can be a sign of a serious heart condition, so detecting heart
murmurs is critical for identifying and managing cardiovascular diseases.
However, current methods for identifying murmurous heart sounds do not fully
utilize the valuable insights that can be gained by exploring intrinsic
properties of heart sound signals. To address this issue, this study proposes a
new discriminatory set of multiscale features based on the self-similarity and
complexity properties of heart sounds, as derived in the wavelet domain.
Self-similarity is characterized by assessing fractal behaviors, while
complexity is explored by calculating wavelet entropy. We evaluated the
diagnostic performance of these proposed features for detecting murmurs using a
set of standard classifiers. When applied to a publicly available heart sound
dataset, our proposed wavelet-based multiscale features achieved comparable
performance to existing methods with fewer features. This suggests that
self-similarity and complexity properties in heart sounds could be potential
biomarkers for improving the accuracy of murmur detection.
| [
{
"version": "v1",
"created": "Sun, 28 May 2023 22:21:31 GMT"
}
] | 2023-06-09T00:00:00 | [
[
"Vimalajeewa",
"Dixon",
""
],
[
"Lee",
"Chihoon",
""
],
[
"Vidakovic",
"Brani",
""
]
] | TITLE: A Method for Detecting Murmurous Heart Sounds based on Self-similar
Properties
ABSTRACT: A heart murmur is an atypical sound produced by the flow of blood through the
heart. It can be a sign of a serious heart condition, so detecting heart
murmurs is critical for identifying and managing cardiovascular diseases.
However, current methods for identifying murmurous heart sounds do not fully
utilize the valuable insights that can be gained by exploring intrinsic
properties of heart sound signals. To address this issue, this study proposes a
new discriminatory set of multiscale features based on the self-similarity and
complexity properties of heart sounds, as derived in the wavelet domain.
Self-similarity is characterized by assessing fractal behaviors, while
complexity is explored by calculating wavelet entropy. We evaluated the
diagnostic performance of these proposed features for detecting murmurs using a
set of standard classifiers. When applied to a publicly available heart sound
dataset, our proposed wavelet-based multiscale features achieved comparable
performance to existing methods with fewer features. This suggests that
self-similarity and complexity properties in heart sounds could be potential
biomarkers for improving the accuracy of murmur detection.
|
2306.05286 | Han Yi Chiu | Han Yi Chiu, Liang Zhao, Anqi Wu | JGAT: a joint spatio-temporal graph attention model for brain decoding | null | null | null | null | q-bio.NC cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The decoding of brain neural networks has been an intriguing topic in
neuroscience for a well-rounded understanding of different types of brain
disorders and cognitive stimuli. Integrating different types of connectivity,
e.g., Functional Connectivity (FC) and Structural Connectivity (SC), from
multi-modal imaging techniques can take their complementary information into
account and therefore have the potential to get better decoding capability.
However, traditional approaches for integrating FC and SC overlook the
dynamical variations, which stand a great chance to over-generalize the brain
neural network. In this paper, we propose a Joint kernel Graph Attention
Network (JGAT), which is a new multi-modal temporal graph attention network
framework. It integrates the data from functional Magnetic Resonance Images
(fMRI) and Diffusion Weighted Imaging (DWI) while preserving the dynamic
information at the same time. We conduct brain-decoding tasks with our JGAT on
four independent datasets: three of 7T fMRI datasets from the Human Connectome
Project (HCP) and one from animal neural recordings. Furthermore, with
Attention Scores (AS) and Frame Scores (FS) computed and learned from the
model, we can locate several informative temporal segments and build meaningful
dynamical pathways along the temporal domain for the HCP datasets. The URL to
the code of JGAT model: https://github.com/BRAINML-GT/JGAT.
| [
{
"version": "v1",
"created": "Sat, 3 Jun 2023 02:45:03 GMT"
}
] | 2023-06-09T00:00:00 | [
[
"Chiu",
"Han Yi",
""
],
[
"Zhao",
"Liang",
""
],
[
"Wu",
"Anqi",
""
]
] | TITLE: JGAT: a joint spatio-temporal graph attention model for brain decoding
ABSTRACT: The decoding of brain neural networks has been an intriguing topic in
neuroscience for a well-rounded understanding of different types of brain
disorders and cognitive stimuli. Integrating different types of connectivity,
e.g., Functional Connectivity (FC) and Structural Connectivity (SC), from
multi-modal imaging techniques can take their complementary information into
account and therefore have the potential to get better decoding capability.
However, traditional approaches for integrating FC and SC overlook the
dynamical variations, which stand a great chance to over-generalize the brain
neural network. In this paper, we propose a Joint kernel Graph Attention
Network (JGAT), which is a new multi-modal temporal graph attention network
framework. It integrates the data from functional Magnetic Resonance Images
(fMRI) and Diffusion Weighted Imaging (DWI) while preserving the dynamic
information at the same time. We conduct brain-decoding tasks with our JGAT on
four independent datasets: three of 7T fMRI datasets from the Human Connectome
Project (HCP) and one from animal neural recordings. Furthermore, with
Attention Scores (AS) and Frame Scores (FS) computed and learned from the
model, we can locate several informative temporal segments and build meaningful
dynamical pathways along the temporal domain for the HCP datasets. The URL to
the code of JGAT model: https://github.com/BRAINML-GT/JGAT.
|
2306.05294 | Massih-Reza Amini | Alkesandra Malkova and Massih-Reza Amini and Benoit Denis and
Christophe Villien | Deep Learning with Partially Labeled Data for Radio Map Reconstruction | 42 pages, 39 figures | null | null | null | eess.SP cs.IT cs.LG math.IT | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this paper, we address the problem of Received Signal Strength map
reconstruction based on location-dependent radio measurements and utilizing
side knowledge about the local region; for example, city plan, terrain height,
gateway position. Depending on the quantity of such prior side information, we
employ Neural Architecture Search to find an optimized Neural Network model
with the best architecture for each of the supposed settings. We demonstrate
that using additional side information enhances the final accuracy of the
Received Signal Strength map reconstruction on three datasets that correspond
to three major cities, particularly in sub-areas near the gateways where larger
variations of the average received signal power are typically observed.
| [
{
"version": "v1",
"created": "Wed, 7 Jun 2023 08:18:56 GMT"
}
] | 2023-06-09T00:00:00 | [
[
"Malkova",
"Alkesandra",
""
],
[
"Amini",
"Massih-Reza",
""
],
[
"Denis",
"Benoit",
""
],
[
"Villien",
"Christophe",
""
]
] | TITLE: Deep Learning with Partially Labeled Data for Radio Map Reconstruction
ABSTRACT: In this paper, we address the problem of Received Signal Strength map
reconstruction based on location-dependent radio measurements and utilizing
side knowledge about the local region; for example, city plan, terrain height,
gateway position. Depending on the quantity of such prior side information, we
employ Neural Architecture Search to find an optimized Neural Network model
with the best architecture for each of the supposed settings. We demonstrate
that using additional side information enhances the final accuracy of the
Received Signal Strength map reconstruction on three datasets that correspond
to three major cities, particularly in sub-areas near the gateways where larger
variations of the average received signal power are typically observed.
|
2306.05297 | Gongshu Wang | Gongshu Wang, Ning Jiang, Yunxiao Ma, Tiantian Liu, Duanduan Chen,
Jinglong Wu, Guoqi Li, Dong Liang, Tianyi Yan | Connectional-Style-Guided Contextual Representation Learning for Brain
Disease Diagnosis | null | null | null | null | eess.IV cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Structural magnetic resonance imaging (sMRI) has shown great clinical value
and has been widely used in deep learning (DL) based computer-aided brain
disease diagnosis. Previous approaches focused on local shapes and textures in
sMRI that may be significant only within a particular domain. The learned
representations are likely to contain spurious information and have a poor
generalization ability in other diseases and datasets. To facilitate capturing
meaningful and robust features, it is necessary to first comprehensively
understand the intrinsic pattern of the brain that is not restricted within a
single data/task domain. Considering that the brain is a complex connectome of
interlinked neurons, the connectional properties in the brain have strong
biological significance, which is shared across multiple domains and covers
most pathological information. In this work, we propose a connectional style
contextual representation learning model (CS-CRL) to capture the intrinsic
pattern of the brain, used for multiple brain disease diagnosis. Specifically,
it has a vision transformer (ViT) encoder and leverages mask reconstruction as
the proxy task and Gram matrices to guide the representation of connectional
information. It facilitates the capture of global context and the aggregation
of features with biological plausibility. The results indicate that CS-CRL
achieves superior accuracy in multiple brain disease diagnosis tasks across six
datasets and three diseases and outperforms state-of-the-art models.
Furthermore, we demonstrate that CS-CRL captures more brain-network-like
properties, better aggregates features, is easier to optimize and is more
robust to noise, which explains its superiority in theory. Our source code will
be released soon.
| [
{
"version": "v1",
"created": "Thu, 8 Jun 2023 15:39:27 GMT"
}
] | 2023-06-09T00:00:00 | [
[
"Wang",
"Gongshu",
""
],
[
"Jiang",
"Ning",
""
],
[
"Ma",
"Yunxiao",
""
],
[
"Liu",
"Tiantian",
""
],
[
"Chen",
"Duanduan",
""
],
[
"Wu",
"Jinglong",
""
],
[
"Li",
"Guoqi",
""
],
[
"Liang",
"Dong",
""
],
[
"Yan",
"Tianyi",
""
]
] | TITLE: Connectional-Style-Guided Contextual Representation Learning for Brain
Disease Diagnosis
ABSTRACT: Structural magnetic resonance imaging (sMRI) has shown great clinical value
and has been widely used in deep learning (DL) based computer-aided brain
disease diagnosis. Previous approaches focused on local shapes and textures in
sMRI that may be significant only within a particular domain. The learned
representations are likely to contain spurious information and have a poor
generalization ability in other diseases and datasets. To facilitate capturing
meaningful and robust features, it is necessary to first comprehensively
understand the intrinsic pattern of the brain that is not restricted within a
single data/task domain. Considering that the brain is a complex connectome of
interlinked neurons, the connectional properties in the brain have strong
biological significance, which is shared across multiple domains and covers
most pathological information. In this work, we propose a connectional style
contextual representation learning model (CS-CRL) to capture the intrinsic
pattern of the brain, used for multiple brain disease diagnosis. Specifically,
it has a vision transformer (ViT) encoder and leverages mask reconstruction as
the proxy task and Gram matrices to guide the representation of connectional
information. It facilitates the capture of global context and the aggregation
of features with biological plausibility. The results indicate that CS-CRL
achieves superior accuracy in multiple brain disease diagnosis tasks across six
datasets and three diseases and outperforms state-of-the-art models.
Furthermore, we demonstrate that CS-CRL captures more brain-network-like
properties, better aggregates features, is easier to optimize and is more
robust to noise, which explains its superiority in theory. Our source code will
be released soon.
|
2306.05307 | Fanny Jourdan | Fanny Jourdan, Laurent Risser, Jean-Michel Loubes, Nicholas Asher | Are fairness metric scores enough to assess discrimination biases in
machine learning? | Accepted for publication at Third Workshop on Trustworthy Natural
Language Processing, ACL 2023 | null | null | null | cs.CL cs.CY cs.LG stat.ML | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | This paper presents novel experiments shedding light on the shortcomings of
current metrics for assessing biases of gender discrimination made by machine
learning algorithms on textual data. We focus on the Bios dataset, and our
learning task is to predict the occupation of individuals, based on their
biography. Such prediction tasks are common in commercial Natural Language
Processing (NLP) applications such as automatic job recommendations. We address
an important limitation of theoretical discussions dealing with group-wise
fairness metrics: they focus on large datasets, although the norm in many
industrial NLP applications is to use small to reasonably large linguistic
datasets for which the main practical constraint is to get a good prediction
accuracy. We then question how reliable are different popular measures of bias
when the size of the training set is simply sufficient to learn reasonably
accurate predictions. Our experiments sample the Bios dataset and learn more
than 200 models on different sample sizes. This allows us to statistically
study our results and to confirm that common gender bias indices provide
diverging and sometimes unreliable results when applied to relatively small
training and test samples. This highlights the crucial importance of variance
calculations for providing sound results in this field.
| [
{
"version": "v1",
"created": "Thu, 8 Jun 2023 15:56:57 GMT"
}
] | 2023-06-09T00:00:00 | [
[
"Jourdan",
"Fanny",
""
],
[
"Risser",
"Laurent",
""
],
[
"Loubes",
"Jean-Michel",
""
],
[
"Asher",
"Nicholas",
""
]
] | TITLE: Are fairness metric scores enough to assess discrimination biases in
machine learning?
ABSTRACT: This paper presents novel experiments shedding light on the shortcomings of
current metrics for assessing biases of gender discrimination made by machine
learning algorithms on textual data. We focus on the Bios dataset, and our
learning task is to predict the occupation of individuals, based on their
biography. Such prediction tasks are common in commercial Natural Language
Processing (NLP) applications such as automatic job recommendations. We address
an important limitation of theoretical discussions dealing with group-wise
fairness metrics: they focus on large datasets, although the norm in many
industrial NLP applications is to use small to reasonably large linguistic
datasets for which the main practical constraint is to get a good prediction
accuracy. We then question how reliable are different popular measures of bias
when the size of the training set is simply sufficient to learn reasonably
accurate predictions. Our experiments sample the Bios dataset and learn more
than 200 models on different sample sizes. This allows us to statistically
study our results and to confirm that common gender bias indices provide
diverging and sometimes unreliable results when applied to relatively small
training and test samples. This highlights the crucial importance of variance
calculations for providing sound results in this field.
|
2306.05331 | Yangyang Yu | Yangyang Yu, Jordan W. Suchow | Actively learning a Bayesian matrix fusion model with deep side
information | null | null | null | null | cs.AI cs.CE | http://creativecommons.org/licenses/by/4.0/ | High-dimensional deep neural network representations of images and concepts
can be aligned to predict human annotations of diverse stimuli. However, such
alignment requires the costly collection of behavioral responses, such that, in
practice, the deep-feature spaces are only ever sparsely sampled. Here, we
propose an active learning approach to adaptively sampling experimental stimuli
to efficiently learn a Bayesian matrix factorization model with deep side
information. We observe a significant efficiency gain over a passive baseline.
Furthermore, with a sequential batched sampling strategy, the algorithm is
applicable not only to small datasets collected from traditional laboratory
experiments but also to settings where large-scale crowdsourced data collection
is needed to accurately align the high-dimensional deep feature representations
derived from pre-trained networks.
| [
{
"version": "v1",
"created": "Thu, 8 Jun 2023 16:31:47 GMT"
}
] | 2023-06-09T00:00:00 | [
[
"Yu",
"Yangyang",
""
],
[
"Suchow",
"Jordan W.",
""
]
] | TITLE: Actively learning a Bayesian matrix fusion model with deep side
information
ABSTRACT: High-dimensional deep neural network representations of images and concepts
can be aligned to predict human annotations of diverse stimuli. However, such
alignment requires the costly collection of behavioral responses, such that, in
practice, the deep-feature spaces are only ever sparsely sampled. Here, we
propose an active learning approach to adaptively sampling experimental stimuli
to efficiently learn a Bayesian matrix factorization model with deep side
information. We observe a significant efficiency gain over a passive baseline.
Furthermore, with a sequential batched sampling strategy, the algorithm is
applicable not only to small datasets collected from traditional laboratory
experiments but also to settings where large-scale crowdsourced data collection
is needed to accurately align the high-dimensional deep feature representations
derived from pre-trained networks.
|
2306.05370 | Palina Niamkova | Poli Nemkova, Solomon Ubani, Suleyman Olcay Polat, Nayeon Kim, Rodney
D. Nielsen | Detecting Human Rights Violations on Social Media during Russia-Ukraine
War | 9 pages | null | null | null | cs.CY cs.CL | http://creativecommons.org/licenses/by-nc-nd/4.0/ | The present-day Russia-Ukraine military conflict has exposed the pivotal role
of social media in enabling the transparent and unbridled sharing of
information directly from the frontlines. In conflict zones where freedom of
expression is constrained and information warfare is pervasive, social media
has emerged as an indispensable lifeline. Anonymous social media platforms, as
publicly available sources for disseminating war-related information, have the
potential to serve as effective instruments for monitoring and documenting
Human Rights Violations (HRV). Our research focuses on the analysis of data
from Telegram, the leading social media platform for reading independent news
in post-Soviet regions. We gathered a dataset of posts sampled from 95 public
Telegram channels that cover politics and war news, which we have utilized to
identify potential occurrences of HRV. Employing a mBERT-based text classifier,
we have conducted an analysis to detect any mentions of HRV in the Telegram
data. Our final approach yielded an $F_2$ score of 0.71 for HRV detection,
representing an improvement of 0.38 over the multilingual BERT base model. We
release two datasets that contains Telegram posts: (1) large corpus with over
2.3 millions posts and (2) annotated at the sentence-level dataset to indicate
HRVs. The Telegram posts are in the context of the Russia-Ukraine war. We posit
that our findings hold significant implications for NGOs, governments, and
researchers by providing a means to detect and document possible human rights
violations.
| [
{
"version": "v1",
"created": "Tue, 6 Jun 2023 12:59:03 GMT"
}
] | 2023-06-09T00:00:00 | [
[
"Nemkova",
"Poli",
""
],
[
"Ubani",
"Solomon",
""
],
[
"Polat",
"Suleyman Olcay",
""
],
[
"Kim",
"Nayeon",
""
],
[
"Nielsen",
"Rodney D.",
""
]
] | TITLE: Detecting Human Rights Violations on Social Media during Russia-Ukraine
War
ABSTRACT: The present-day Russia-Ukraine military conflict has exposed the pivotal role
of social media in enabling the transparent and unbridled sharing of
information directly from the frontlines. In conflict zones where freedom of
expression is constrained and information warfare is pervasive, social media
has emerged as an indispensable lifeline. Anonymous social media platforms, as
publicly available sources for disseminating war-related information, have the
potential to serve as effective instruments for monitoring and documenting
Human Rights Violations (HRV). Our research focuses on the analysis of data
from Telegram, the leading social media platform for reading independent news
in post-Soviet regions. We gathered a dataset of posts sampled from 95 public
Telegram channels that cover politics and war news, which we have utilized to
identify potential occurrences of HRV. Employing a mBERT-based text classifier,
we have conducted an analysis to detect any mentions of HRV in the Telegram
data. Our final approach yielded an $F_2$ score of 0.71 for HRV detection,
representing an improvement of 0.38 over the multilingual BERT base model. We
release two datasets that contains Telegram posts: (1) large corpus with over
2.3 millions posts and (2) annotated at the sentence-level dataset to indicate
HRVs. The Telegram posts are in the context of the Russia-Ukraine war. We posit
that our findings hold significant implications for NGOs, governments, and
researchers by providing a means to detect and document possible human rights
violations.
|
2306.05375 | Anwar Said | Ammar Ahmed, Anwar Said, Mudassir Shabbir, Xenofon Koutsoukos | Sequential Graph Neural Networks for Source Code Vulnerability
Identification | 7 pages paper presented at HotSoS2023 conference | null | null | null | cs.CR cs.LG | http://creativecommons.org/licenses/by/4.0/ | Vulnerability identification constitutes a task of high importance for cyber
security. It is quite helpful for locating and fixing vulnerable functions in
large applications. However, this task is rather challenging owing to the
absence of reliable and adequately managed datasets and learning models.
Existing solutions typically rely on human expertise to annotate datasets or
specify features, which is prone to error. In addition, the learning models
have a high rate of false positives. To bridge this gap, in this paper, we
present a properly curated C/C++ source code vulnerability dataset, denoted as
CVEFunctionGraphEmbeddings (CVEFGE), to aid in developing models. CVEFGE is
automatically crawled from the CVE database, which contains authentic and
publicly disclosed source code vulnerabilities. We also propose a learning
framework based on graph neural networks, denoted SEquential Graph Neural
Network (SEGNN) for learning a large number of code semantic representations.
SEGNN consists of a sequential learning module, graph convolution, pooling, and
fully connected layers. Our evaluations on two datasets and four baseline
methods in a graph classification setting demonstrate state-of-the-art results.
| [
{
"version": "v1",
"created": "Tue, 23 May 2023 17:25:51 GMT"
}
] | 2023-06-09T00:00:00 | [
[
"Ahmed",
"Ammar",
""
],
[
"Said",
"Anwar",
""
],
[
"Shabbir",
"Mudassir",
""
],
[
"Koutsoukos",
"Xenofon",
""
]
] | TITLE: Sequential Graph Neural Networks for Source Code Vulnerability
Identification
ABSTRACT: Vulnerability identification constitutes a task of high importance for cyber
security. It is quite helpful for locating and fixing vulnerable functions in
large applications. However, this task is rather challenging owing to the
absence of reliable and adequately managed datasets and learning models.
Existing solutions typically rely on human expertise to annotate datasets or
specify features, which is prone to error. In addition, the learning models
have a high rate of false positives. To bridge this gap, in this paper, we
present a properly curated C/C++ source code vulnerability dataset, denoted as
CVEFunctionGraphEmbeddings (CVEFGE), to aid in developing models. CVEFGE is
automatically crawled from the CVE database, which contains authentic and
publicly disclosed source code vulnerabilities. We also propose a learning
framework based on graph neural networks, denoted SEquential Graph Neural
Network (SEGNN) for learning a large number of code semantic representations.
SEGNN consists of a sequential learning module, graph convolution, pooling, and
fully connected layers. Our evaluations on two datasets and four baseline
methods in a graph classification setting demonstrate state-of-the-art results.
|
2306.05376 | Akash Awasthi | Akash Awasthi, Son Ly, Jaer Nizam, Samira Zare, Videet Mehta, Safwan
Ahmed, Keshav Shah, Ramakrishna Nemani, Saurabh Prasad, Hien Van Nguyen | Anomaly Detection in Satellite Videos using Diffusion Models | null | null | null | null | cs.CV cs.LG | http://creativecommons.org/licenses/by/4.0/ | The definition of anomaly detection is the identification of an unexpected
event. Real-time detection of extreme events such as wildfires, cyclones, or
floods using satellite data has become crucial for disaster management.
Although several earth-observing satellites provide information about
disasters, satellites in the geostationary orbit provide data at intervals as
frequent as every minute, effectively creating a video from space. There are
many techniques that have been proposed to identify anomalies in surveillance
videos; however, the available datasets do not have dynamic behavior, so we
discuss an anomaly framework that can work on very high-frequency datasets to
find very fast-moving anomalies. In this work, we present a diffusion model
which does not need any motion component to capture the fast-moving anomalies
and outperforms the other baseline methods.
| [
{
"version": "v1",
"created": "Thu, 25 May 2023 19:17:39 GMT"
}
] | 2023-06-09T00:00:00 | [
[
"Awasthi",
"Akash",
""
],
[
"Ly",
"Son",
""
],
[
"Nizam",
"Jaer",
""
],
[
"Zare",
"Samira",
""
],
[
"Mehta",
"Videet",
""
],
[
"Ahmed",
"Safwan",
""
],
[
"Shah",
"Keshav",
""
],
[
"Nemani",
"Ramakrishna",
""
],
[
"Prasad",
"Saurabh",
""
],
[
"Van Nguyen",
"Hien",
""
]
] | TITLE: Anomaly Detection in Satellite Videos using Diffusion Models
ABSTRACT: The definition of anomaly detection is the identification of an unexpected
event. Real-time detection of extreme events such as wildfires, cyclones, or
floods using satellite data has become crucial for disaster management.
Although several earth-observing satellites provide information about
disasters, satellites in the geostationary orbit provide data at intervals as
frequent as every minute, effectively creating a video from space. There are
many techniques that have been proposed to identify anomalies in surveillance
videos; however, the available datasets do not have dynamic behavior, so we
discuss an anomaly framework that can work on very high-frequency datasets to
find very fast-moving anomalies. In this work, we present a diffusion model
which does not need any motion component to capture the fast-moving anomalies
and outperforms the other baseline methods.
|
2306.05390 | Dongdong Chen | Qinhong Yang and Dongdong Chen and Zhentao Tan and Qiankun Liu and Qi
Chu and Jianmin Bao and Lu Yuan and Gang Hua and Nenghai Yu | HQ-50K: A Large-scale, High-quality Dataset for Image Restoration | Dataset and code will be available at
https://github.com/littleYaang/HQ-50K | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | This paper introduces a new large-scale image restoration dataset, called
HQ-50K, which contains 50,000 high-quality images with rich texture details and
semantic diversity. We analyze existing image restoration datasets from five
different perspectives, including data scale, resolution, compression rates,
texture details, and semantic coverage. However, we find that all of these
datasets are deficient in some aspects. In contrast, HQ-50K considers all of
these five aspects during the data curation process and meets all requirements.
We also present a new Degradation-Aware Mixture of Expert (DAMoE) model, which
enables a single model to handle multiple corruption types and unknown levels.
Our extensive experiments demonstrate that HQ-50K consistently improves the
performance on various image restoration tasks, such as super-resolution,
denoising, dejpeg, and deraining. Furthermore, our proposed DAMoE, trained on
our \dataset, outperforms existing state-of-the-art unified models designed for
multiple restoration tasks and levels. The dataset and code are available at
\url{https://github.com/littleYaang/HQ-50K}.
| [
{
"version": "v1",
"created": "Thu, 8 Jun 2023 17:44:21 GMT"
}
] | 2023-06-09T00:00:00 | [
[
"Yang",
"Qinhong",
""
],
[
"Chen",
"Dongdong",
""
],
[
"Tan",
"Zhentao",
""
],
[
"Liu",
"Qiankun",
""
],
[
"Chu",
"Qi",
""
],
[
"Bao",
"Jianmin",
""
],
[
"Yuan",
"Lu",
""
],
[
"Hua",
"Gang",
""
],
[
"Yu",
"Nenghai",
""
]
] | TITLE: HQ-50K: A Large-scale, High-quality Dataset for Image Restoration
ABSTRACT: This paper introduces a new large-scale image restoration dataset, called
HQ-50K, which contains 50,000 high-quality images with rich texture details and
semantic diversity. We analyze existing image restoration datasets from five
different perspectives, including data scale, resolution, compression rates,
texture details, and semantic coverage. However, we find that all of these
datasets are deficient in some aspects. In contrast, HQ-50K considers all of
these five aspects during the data curation process and meets all requirements.
We also present a new Degradation-Aware Mixture of Expert (DAMoE) model, which
enables a single model to handle multiple corruption types and unknown levels.
Our extensive experiments demonstrate that HQ-50K consistently improves the
performance on various image restoration tasks, such as super-resolution,
denoising, dejpeg, and deraining. Furthermore, our proposed DAMoE, trained on
our \dataset, outperforms existing state-of-the-art unified models designed for
multiple restoration tasks and levels. The dataset and code are available at
\url{https://github.com/littleYaang/HQ-50K}.
|
2306.05392 | Sanjay Subramanian | Sanjay Subramanian, Medhini Narasimhan, Kushal Khangaonkar, Kevin
Yang, Arsha Nagrani, Cordelia Schmid, Andy Zeng, Trevor Darrell, Dan Klein | Modular Visual Question Answering via Code Generation | ACL 2023 | null | null | null | cs.CL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We present a framework that formulates visual question answering as modular
code generation. In contrast to prior work on modular approaches to VQA, our
approach requires no additional training and relies on pre-trained language
models (LMs), visual models pre-trained on image-caption pairs, and fifty VQA
examples used for in-context learning. The generated Python programs invoke and
compose the outputs of the visual models using arithmetic and conditional
logic. Our approach improves accuracy on the COVR dataset by at least 3% and on
the GQA dataset by roughly 2% compared to the few-shot baseline that does not
employ code generation.
| [
{
"version": "v1",
"created": "Thu, 8 Jun 2023 17:45:14 GMT"
}
] | 2023-06-09T00:00:00 | [
[
"Subramanian",
"Sanjay",
""
],
[
"Narasimhan",
"Medhini",
""
],
[
"Khangaonkar",
"Kushal",
""
],
[
"Yang",
"Kevin",
""
],
[
"Nagrani",
"Arsha",
""
],
[
"Schmid",
"Cordelia",
""
],
[
"Zeng",
"Andy",
""
],
[
"Darrell",
"Trevor",
""
],
[
"Klein",
"Dan",
""
]
] | TITLE: Modular Visual Question Answering via Code Generation
ABSTRACT: We present a framework that formulates visual question answering as modular
code generation. In contrast to prior work on modular approaches to VQA, our
approach requires no additional training and relies on pre-trained language
models (LMs), visual models pre-trained on image-caption pairs, and fifty VQA
examples used for in-context learning. The generated Python programs invoke and
compose the outputs of the visual models using arithmetic and conditional
logic. Our approach improves accuracy on the COVR dataset by at least 3% and on
the GQA dataset by roughly 2% compared to the few-shot baseline that does not
employ code generation.
|
2306.05419 | M. Esat Kalfaoglu | M. Esat Kalfaoglu, Halil Ibrahim Ozturk, Ozsel Kilinc, Alptekin
Temizel | TopoMask: Instance-Mask-Based Formulation for the Road Topology Problem
via Transformer-Based Architecture | 4th in OLS and 2nd in the F1-score in OpenLane Topology Challenge
2023 | null | null | null | cs.CV cs.AI cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Driving scene understanding task involves detecting static elements such as
lanes, traffic signs, and traffic lights, and their relationships with each
other. To facilitate the development of comprehensive scene understanding
solutions using multiple camera views, a new dataset called Road Genome
(OpenLane-V2) has been released. This dataset allows for the exploration of
complex road connections and situations where lane markings may be absent.
Instead of using traditional lane markings, the lanes in this dataset are
represented by centerlines, which offer a more suitable representation of lanes
and their connections. In this study, we have introduced a new approach called
TopoMask for predicting centerlines in road topology. Unlike existing
approaches in the literature that rely on keypoints or parametric methods,
TopoMask utilizes an instance-mask based formulation with a transformer-based
architecture and, in order to enrich the mask instances with flow information,
a direction label representation is proposed. TopoMask have ranked 4th in the
OpenLane-V2 Score (OLS) and ranked 2nd in the F1 score of centerline prediction
in OpenLane Topology Challenge 2023. In comparison to the current
state-of-the-art method, TopoNet, the proposed method has achieved similar
performance in Frechet-based lane detection and outperformed TopoNet in
Chamfer-based lane detection without utilizing its scene graph neural network.
| [
{
"version": "v1",
"created": "Thu, 8 Jun 2023 17:58:57 GMT"
}
] | 2023-06-09T00:00:00 | [
[
"Kalfaoglu",
"M. Esat",
""
],
[
"Ozturk",
"Halil Ibrahim",
""
],
[
"Kilinc",
"Ozsel",
""
],
[
"Temizel",
"Alptekin",
""
]
] | TITLE: TopoMask: Instance-Mask-Based Formulation for the Road Topology Problem
via Transformer-Based Architecture
ABSTRACT: Driving scene understanding task involves detecting static elements such as
lanes, traffic signs, and traffic lights, and their relationships with each
other. To facilitate the development of comprehensive scene understanding
solutions using multiple camera views, a new dataset called Road Genome
(OpenLane-V2) has been released. This dataset allows for the exploration of
complex road connections and situations where lane markings may be absent.
Instead of using traditional lane markings, the lanes in this dataset are
represented by centerlines, which offer a more suitable representation of lanes
and their connections. In this study, we have introduced a new approach called
TopoMask for predicting centerlines in road topology. Unlike existing
approaches in the literature that rely on keypoints or parametric methods,
TopoMask utilizes an instance-mask based formulation with a transformer-based
architecture and, in order to enrich the mask instances with flow information,
a direction label representation is proposed. TopoMask have ranked 4th in the
OpenLane-V2 Score (OLS) and ranked 2nd in the F1 score of centerline prediction
in OpenLane Topology Challenge 2023. In comparison to the current
state-of-the-art method, TopoNet, the proposed method has achieved similar
performance in Frechet-based lane detection and outperformed TopoNet in
Chamfer-based lane detection without utilizing its scene graph neural network.
|
2306.05420 | Carlos Esteves | Carlos Esteves, Jean-Jacques Slotine, Ameesh Makadia | Scaling Spherical CNNs | Accepted to ICML'23 | null | null | null | cs.LG cs.CV | http://creativecommons.org/licenses/by/4.0/ | Spherical CNNs generalize CNNs to functions on the sphere, by using spherical
convolutions as the main linear operation. The most accurate and efficient way
to compute spherical convolutions is in the spectral domain (via the
convolution theorem), which is still costlier than the usual planar
convolutions. For this reason, applications of spherical CNNs have so far been
limited to small problems that can be approached with low model capacity. In
this work, we show how spherical CNNs can be scaled for much larger problems.
To achieve this, we make critical improvements including novel variants of
common model components, an implementation of core operations to exploit
hardware accelerator characteristics, and application-specific input
representations that exploit the properties of our model. Experiments show our
larger spherical CNNs reach state-of-the-art on several targets of the QM9
molecular benchmark, which was previously dominated by equivariant graph neural
networks, and achieve competitive performance on multiple weather forecasting
tasks. Our code is available at
https://github.com/google-research/spherical-cnn.
| [
{
"version": "v1",
"created": "Thu, 8 Jun 2023 17:59:08 GMT"
}
] | 2023-06-09T00:00:00 | [
[
"Esteves",
"Carlos",
""
],
[
"Slotine",
"Jean-Jacques",
""
],
[
"Makadia",
"Ameesh",
""
]
] | TITLE: Scaling Spherical CNNs
ABSTRACT: Spherical CNNs generalize CNNs to functions on the sphere, by using spherical
convolutions as the main linear operation. The most accurate and efficient way
to compute spherical convolutions is in the spectral domain (via the
convolution theorem), which is still costlier than the usual planar
convolutions. For this reason, applications of spherical CNNs have so far been
limited to small problems that can be approached with low model capacity. In
this work, we show how spherical CNNs can be scaled for much larger problems.
To achieve this, we make critical improvements including novel variants of
common model components, an implementation of core operations to exploit
hardware accelerator characteristics, and application-specific input
representations that exploit the properties of our model. Experiments show our
larger spherical CNNs reach state-of-the-art on several targets of the QM9
molecular benchmark, which was previously dominated by equivariant graph neural
networks, and achieve competitive performance on multiple weather forecasting
tasks. Our code is available at
https://github.com/google-research/spherical-cnn.
|
2306.05421 | Sirui Xu | Sirui Xu, Yu-Xiong Wang, Liang-Yan Gui | Stochastic Multi-Person 3D Motion Forecasting | ICLR 2023 (Top 25% Paper); Project Page:
https://sirui-xu.github.io/DuMMF | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | This paper aims to deal with the ignored real-world complexities in prior
work on human motion forecasting, emphasizing the social properties of
multi-person motion, the diversity of motion and social interactions, and the
complexity of articulated motion. To this end, we introduce a novel task of
stochastic multi-person 3D motion forecasting. We propose a dual-level
generative modeling framework that separately models independent individual
motion at the local level and social interactions at the global level. Notably,
this dual-level modeling mechanism can be achieved within a shared generative
model, through introducing learnable latent codes that represent intents of
future motion and switching the codes' modes of operation at different levels.
Our framework is general; we instantiate it with different generative models,
including generative adversarial networks and diffusion models, and various
multi-person forecasting models. Extensive experiments on CMU-Mocap, MuPoTS-3D,
and SoMoF benchmarks show that our approach produces diverse and accurate
multi-person predictions, significantly outperforming the state of the art.
| [
{
"version": "v1",
"created": "Thu, 8 Jun 2023 17:59:09 GMT"
}
] | 2023-06-09T00:00:00 | [
[
"Xu",
"Sirui",
""
],
[
"Wang",
"Yu-Xiong",
""
],
[
"Gui",
"Liang-Yan",
""
]
] | TITLE: Stochastic Multi-Person 3D Motion Forecasting
ABSTRACT: This paper aims to deal with the ignored real-world complexities in prior
work on human motion forecasting, emphasizing the social properties of
multi-person motion, the diversity of motion and social interactions, and the
complexity of articulated motion. To this end, we introduce a novel task of
stochastic multi-person 3D motion forecasting. We propose a dual-level
generative modeling framework that separately models independent individual
motion at the local level and social interactions at the global level. Notably,
this dual-level modeling mechanism can be achieved within a shared generative
model, through introducing learnable latent codes that represent intents of
future motion and switching the codes' modes of operation at different levels.
Our framework is general; we instantiate it with different generative models,
including generative adversarial networks and diffusion models, and various
multi-person forecasting models. Extensive experiments on CMU-Mocap, MuPoTS-3D,
and SoMoF benchmarks show that our approach produces diverse and accurate
multi-person predictions, significantly outperforming the state of the art.
|
2306.05425 | Liangyu Chen | Bo Li, Yuanhan Zhang, Liangyu Chen, Jinghao Wang, Fanyi Pu, Jingkang
Yang, Chunyuan Li, Ziwei Liu | MIMIC-IT: Multi-Modal In-Context Instruction Tuning | Project page: https://otter-ntu.github.io/ Dataset & code:
https://github.com/Luodian/otter Initial release, work in progress | null | null | null | cs.CV cs.AI cs.CL cs.HC | http://creativecommons.org/licenses/by-nc-nd/4.0/ | High-quality instructions and responses are essential for the zero-shot
performance of large language models on interactive natural language tasks. For
interactive vision-language tasks involving intricate visual scenes, a large
quantity of diverse and creative instruction-response pairs should be
imperative to tune vision-language models (VLMs). Nevertheless, the current
availability of vision-language instruction-response pairs in terms of
quantity, diversity, and creativity remains limited, posing challenges to the
generalization of interactive VLMs. Here we present MultI-Modal In-Context
Instruction Tuning (MIMIC-IT), a dataset comprising 2.8 million multimodal
instruction-response pairs, with 2.2 million unique instructions derived from
images and videos. Each pair is accompanied by multi-modal in-context
information, forming conversational contexts aimed at empowering VLMs in
perception, reasoning, and planning. The instruction-response collection
process, dubbed as Syphus, is scaled using an automatic annotation pipeline
that combines human expertise with GPT's capabilities. Using the MIMIC-IT
dataset, we train a large VLM named Otter. Based on extensive evaluations
conducted on vision-language benchmarks, it has been observed that Otter
demonstrates remarkable proficiency in multi-modal perception, reasoning, and
in-context learning. Human evaluation reveals it effectively aligns with the
user's intentions. We release the MIMIC-IT dataset, instruction-response
collection pipeline, benchmarks, and the Otter model.
| [
{
"version": "v1",
"created": "Thu, 8 Jun 2023 17:59:56 GMT"
}
] | 2023-06-09T00:00:00 | [
[
"Li",
"Bo",
""
],
[
"Zhang",
"Yuanhan",
""
],
[
"Chen",
"Liangyu",
""
],
[
"Wang",
"Jinghao",
""
],
[
"Pu",
"Fanyi",
""
],
[
"Yang",
"Jingkang",
""
],
[
"Li",
"Chunyuan",
""
],
[
"Liu",
"Ziwei",
""
]
] | TITLE: MIMIC-IT: Multi-Modal In-Context Instruction Tuning
ABSTRACT: High-quality instructions and responses are essential for the zero-shot
performance of large language models on interactive natural language tasks. For
interactive vision-language tasks involving intricate visual scenes, a large
quantity of diverse and creative instruction-response pairs should be
imperative to tune vision-language models (VLMs). Nevertheless, the current
availability of vision-language instruction-response pairs in terms of
quantity, diversity, and creativity remains limited, posing challenges to the
generalization of interactive VLMs. Here we present MultI-Modal In-Context
Instruction Tuning (MIMIC-IT), a dataset comprising 2.8 million multimodal
instruction-response pairs, with 2.2 million unique instructions derived from
images and videos. Each pair is accompanied by multi-modal in-context
information, forming conversational contexts aimed at empowering VLMs in
perception, reasoning, and planning. The instruction-response collection
process, dubbed as Syphus, is scaled using an automatic annotation pipeline
that combines human expertise with GPT's capabilities. Using the MIMIC-IT
dataset, we train a large VLM named Otter. Based on extensive evaluations
conducted on vision-language benchmarks, it has been observed that Otter
demonstrates remarkable proficiency in multi-modal perception, reasoning, and
in-context learning. Human evaluation reveals it effectively aligns with the
user's intentions. We release the MIMIC-IT dataset, instruction-response
collection pipeline, benchmarks, and the Otter model.
|
2306.05428 | Manel Baradad Jurjo | Manel Baradad, Yuanzhen Li, Forrester Cole, Michael Rubinstein,
Antonio Torralba, William T. Freeman, Varun Jampani | Background Prompting for Improved Object Depth | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Estimating the depth of objects from a single image is a valuable task for
many vision, robotics, and graphics applications. However, current methods
often fail to produce accurate depth for objects in diverse scenes. In this
work, we propose a simple yet effective Background Prompting strategy that
adapts the input object image with a learned background. We learn the
background prompts only using small-scale synthetic object datasets. To infer
object depth on a real image, we place the segmented object into the learned
background prompt and run off-the-shelf depth networks. Background Prompting
helps the depth networks focus on the foreground object, as they are made
invariant to background variations. Moreover, Background Prompting minimizes
the domain gap between synthetic and real object images, leading to better
sim2real generalization than simple finetuning. Results on multiple synthetic
and real datasets demonstrate consistent improvements in real object depths for
a variety of existing depth networks. Code and optimized background prompts can
be found at: https://mbaradad.github.io/depth_prompt.
| [
{
"version": "v1",
"created": "Thu, 8 Jun 2023 17:59:59 GMT"
}
] | 2023-06-09T00:00:00 | [
[
"Baradad",
"Manel",
""
],
[
"Li",
"Yuanzhen",
""
],
[
"Cole",
"Forrester",
""
],
[
"Rubinstein",
"Michael",
""
],
[
"Torralba",
"Antonio",
""
],
[
"Freeman",
"William T.",
""
],
[
"Jampani",
"Varun",
""
]
] | TITLE: Background Prompting for Improved Object Depth
ABSTRACT: Estimating the depth of objects from a single image is a valuable task for
many vision, robotics, and graphics applications. However, current methods
often fail to produce accurate depth for objects in diverse scenes. In this
work, we propose a simple yet effective Background Prompting strategy that
adapts the input object image with a learned background. We learn the
background prompts only using small-scale synthetic object datasets. To infer
object depth on a real image, we place the segmented object into the learned
background prompt and run off-the-shelf depth networks. Background Prompting
helps the depth networks focus on the foreground object, as they are made
invariant to background variations. Moreover, Background Prompting minimizes
the domain gap between synthetic and real object images, leading to better
sim2real generalization than simple finetuning. Results on multiple synthetic
and real datasets demonstrate consistent improvements in real object depths for
a variety of existing depth networks. Code and optimized background prompts can
be found at: https://mbaradad.github.io/depth_prompt.
|
2206.05260 | Emanuel S\'anchez Aimar | Emanuel Sanchez Aimar, Arvi Jonnarth, Michael Felsberg, Marco Kuhlmann | Balanced Product of Calibrated Experts for Long-Tailed Recognition | Accepted at CVPR 2023, 19 pages | null | null | null | cs.CV cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Many real-world recognition problems are characterized by long-tailed label
distributions. These distributions make representation learning highly
challenging due to limited generalization over the tail classes. If the test
distribution differs from the training distribution, e.g. uniform versus
long-tailed, the problem of the distribution shift needs to be addressed. A
recent line of work proposes learning multiple diverse experts to tackle this
issue. Ensemble diversity is encouraged by various techniques, e.g. by
specializing different experts in the head and the tail classes. In this work,
we take an analytical approach and extend the notion of logit adjustment to
ensembles to form a Balanced Product of Experts (BalPoE). BalPoE combines a
family of experts with different test-time target distributions, generalizing
several previous approaches. We show how to properly define these distributions
and combine the experts in order to achieve unbiased predictions, by proving
that the ensemble is Fisher-consistent for minimizing the balanced error. Our
theoretical analysis shows that our balanced ensemble requires calibrated
experts, which we achieve in practice using mixup. We conduct extensive
experiments and our method obtains new state-of-the-art results on three
long-tailed datasets: CIFAR-100-LT, ImageNet-LT, and iNaturalist-2018. Our code
is available at https://github.com/emasa/BalPoE-CalibratedLT.
| [
{
"version": "v1",
"created": "Fri, 10 Jun 2022 17:59:02 GMT"
},
{
"version": "v2",
"created": "Thu, 24 Nov 2022 12:14:59 GMT"
},
{
"version": "v3",
"created": "Wed, 7 Jun 2023 17:52:01 GMT"
}
] | 2023-06-08T00:00:00 | [
[
"Aimar",
"Emanuel Sanchez",
""
],
[
"Jonnarth",
"Arvi",
""
],
[
"Felsberg",
"Michael",
""
],
[
"Kuhlmann",
"Marco",
""
]
] | TITLE: Balanced Product of Calibrated Experts for Long-Tailed Recognition
ABSTRACT: Many real-world recognition problems are characterized by long-tailed label
distributions. These distributions make representation learning highly
challenging due to limited generalization over the tail classes. If the test
distribution differs from the training distribution, e.g. uniform versus
long-tailed, the problem of the distribution shift needs to be addressed. A
recent line of work proposes learning multiple diverse experts to tackle this
issue. Ensemble diversity is encouraged by various techniques, e.g. by
specializing different experts in the head and the tail classes. In this work,
we take an analytical approach and extend the notion of logit adjustment to
ensembles to form a Balanced Product of Experts (BalPoE). BalPoE combines a
family of experts with different test-time target distributions, generalizing
several previous approaches. We show how to properly define these distributions
and combine the experts in order to achieve unbiased predictions, by proving
that the ensemble is Fisher-consistent for minimizing the balanced error. Our
theoretical analysis shows that our balanced ensemble requires calibrated
experts, which we achieve in practice using mixup. We conduct extensive
experiments and our method obtains new state-of-the-art results on three
long-tailed datasets: CIFAR-100-LT, ImageNet-LT, and iNaturalist-2018. Our code
is available at https://github.com/emasa/BalPoE-CalibratedLT.
|
2207.02434 | Mohsen Hooshmand | Shahin Yousefi, Mohsen Hooshmand, Mohsen Afsharchi | Early Discovery of Emerging Entities in Persian Twitter with Semantic
Similarity | null | null | 10.1109/ICWR57742.2023.10139028 | null | cs.CL cs.LG cs.SI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Discovering emerging entities (EEs) is the problem of finding entities before
their establishment. These entities can be critical for individuals, companies,
and governments. Many of these entities can be discovered on social media
platforms, e.g. Twitter. These identities have been the spot of research in
academia and industry in recent years. Similar to any machine learning problem,
data availability is one of the major challenges in this problem. This paper
proposes EEPT. That is an online clustering method able to discover EEs without
any need for training on a dataset. Additionally, due to the lack of a proper
evaluation metric, this paper uses a new metric to evaluate the results. The
results show that EEPT is promising and finds significant entities before their
establishment.
| [
{
"version": "v1",
"created": "Wed, 6 Jul 2022 04:46:07 GMT"
},
{
"version": "v2",
"created": "Wed, 7 Jun 2023 04:51:48 GMT"
}
] | 2023-06-08T00:00:00 | [
[
"Yousefi",
"Shahin",
""
],
[
"Hooshmand",
"Mohsen",
""
],
[
"Afsharchi",
"Mohsen",
""
]
] | TITLE: Early Discovery of Emerging Entities in Persian Twitter with Semantic
Similarity
ABSTRACT: Discovering emerging entities (EEs) is the problem of finding entities before
their establishment. These entities can be critical for individuals, companies,
and governments. Many of these entities can be discovered on social media
platforms, e.g. Twitter. These identities have been the spot of research in
academia and industry in recent years. Similar to any machine learning problem,
data availability is one of the major challenges in this problem. This paper
proposes EEPT. That is an online clustering method able to discover EEs without
any need for training on a dataset. Additionally, due to the lack of a proper
evaluation metric, this paper uses a new metric to evaluate the results. The
results show that EEPT is promising and finds significant entities before their
establishment.
|
2207.04600 | Dong Xia | Zhongyuan Lyu and Dong Xia | Optimal Clustering by Lloyd Algorithm for Low-Rank Mixture Model | null | null | null | null | math.ST cs.IT cs.LG math.IT stat.ME stat.TH | http://creativecommons.org/licenses/by/4.0/ | This paper investigates the computational and statistical limits in
clustering matrix-valued observations. We propose a low-rank mixture model
(LrMM), adapted from the classical Gaussian mixture model (GMM) to treat
matrix-valued observations, which assumes low-rankness for population center
matrices. A computationally efficient clustering method is designed by
integrating Lloyd's algorithm and low-rank approximation. Once
well-initialized, the algorithm converges fast and achieves an exponential-type
clustering error rate that is minimax optimal. Meanwhile, we show that a
tensor-based spectral method delivers a good initial clustering. Comparable to
GMM, the minimax optimal clustering error rate is decided by the separation
strength, i.e., the minimal distance between population center matrices. By
exploiting low-rankness, the proposed algorithm is blessed with a weaker
requirement on the separation strength. Unlike GMM, however, the computational
difficulty of LrMM is characterized by the signal strength, i.e., the smallest
non-zero singular values of population center matrices. Evidence is provided
showing that no polynomial-time algorithm is consistent if the signal strength
is not strong enough, even though the separation strength is strong. Intriguing
differences between estimation and clustering under LrMM are discussed. The
merits of low-rank Lloyd's algorithm are confirmed by comprehensive simulation
experiments. Finally, our method outperforms others in the literature on
real-world datasets.
| [
{
"version": "v1",
"created": "Mon, 11 Jul 2022 03:16:10 GMT"
},
{
"version": "v2",
"created": "Wed, 7 Jun 2023 03:25:39 GMT"
}
] | 2023-06-08T00:00:00 | [
[
"Lyu",
"Zhongyuan",
""
],
[
"Xia",
"Dong",
""
]
] | TITLE: Optimal Clustering by Lloyd Algorithm for Low-Rank Mixture Model
ABSTRACT: This paper investigates the computational and statistical limits in
clustering matrix-valued observations. We propose a low-rank mixture model
(LrMM), adapted from the classical Gaussian mixture model (GMM) to treat
matrix-valued observations, which assumes low-rankness for population center
matrices. A computationally efficient clustering method is designed by
integrating Lloyd's algorithm and low-rank approximation. Once
well-initialized, the algorithm converges fast and achieves an exponential-type
clustering error rate that is minimax optimal. Meanwhile, we show that a
tensor-based spectral method delivers a good initial clustering. Comparable to
GMM, the minimax optimal clustering error rate is decided by the separation
strength, i.e., the minimal distance between population center matrices. By
exploiting low-rankness, the proposed algorithm is blessed with a weaker
requirement on the separation strength. Unlike GMM, however, the computational
difficulty of LrMM is characterized by the signal strength, i.e., the smallest
non-zero singular values of population center matrices. Evidence is provided
showing that no polynomial-time algorithm is consistent if the signal strength
is not strong enough, even though the separation strength is strong. Intriguing
differences between estimation and clustering under LrMM are discussed. The
merits of low-rank Lloyd's algorithm are confirmed by comprehensive simulation
experiments. Finally, our method outperforms others in the literature on
real-world datasets.
|
2207.05258 | Zhao Zhuang | Min Zhang and Zhuang Zhao | A fifth-order finite difference HWENO scheme combined with limiter for
hyperbolic conservation laws | 24 pages, 24 figures | Journal of Computational Physics, 472 (2023) 111676 | 10.1016/j.jcp.2022.111676 | null | math.NA cs.NA | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this paper, a simple fifth-order finite difference Hermite WENO (HWENO)
scheme combined with limiter is proposed for one- and two- dimensional
hyperbolic conservation laws. The fluxes in the governing equation are
approximated by the nonlinear HWENO reconstruction which is the combination of
a quintic polynomial with two quadratic polynomials, where the linear weights
can be artificial positive numbers only if the sum equals one. And other fluxes
in the derivative equations are approximated by high-degree polynomials
directly. For the purpose of controlling spurious oscillations, an HWENO
limiter is applied to modify the derivatives. Instead of using the modified
derivatives both in fluxes reconstruction and time discretization as in the
modified HWENO scheme (J. Sci. Comput., 85:29, 2020), we only apply the
modified derivatives in time discretization while remaining the original
derivatives in fluxes reconstruction. Comparing with the modified HWENO scheme,
the proposed HWENO scheme is simpler, more accurate, efficient and higher
resolution. In addition, the HWENO scheme has a more compact spatial
reconstructed stencil and greater efficiency than the classical fifth-order
finite difference WENO scheme of Jiang and Shu. Various benchmark numerical
examples are presented to show the fifth-order accuracy, great efficiency, high
resolution and robustness of the proposed HWENO scheme.
| [
{
"version": "v1",
"created": "Tue, 12 Jul 2022 02:07:17 GMT"
}
] | 2023-06-08T00:00:00 | [
[
"Zhang",
"Min",
""
],
[
"Zhao",
"Zhuang",
""
]
] | TITLE: A fifth-order finite difference HWENO scheme combined with limiter for
hyperbolic conservation laws
ABSTRACT: In this paper, a simple fifth-order finite difference Hermite WENO (HWENO)
scheme combined with limiter is proposed for one- and two- dimensional
hyperbolic conservation laws. The fluxes in the governing equation are
approximated by the nonlinear HWENO reconstruction which is the combination of
a quintic polynomial with two quadratic polynomials, where the linear weights
can be artificial positive numbers only if the sum equals one. And other fluxes
in the derivative equations are approximated by high-degree polynomials
directly. For the purpose of controlling spurious oscillations, an HWENO
limiter is applied to modify the derivatives. Instead of using the modified
derivatives both in fluxes reconstruction and time discretization as in the
modified HWENO scheme (J. Sci. Comput., 85:29, 2020), we only apply the
modified derivatives in time discretization while remaining the original
derivatives in fluxes reconstruction. Comparing with the modified HWENO scheme,
the proposed HWENO scheme is simpler, more accurate, efficient and higher
resolution. In addition, the HWENO scheme has a more compact spatial
reconstructed stencil and greater efficiency than the classical fifth-order
finite difference WENO scheme of Jiang and Shu. Various benchmark numerical
examples are presented to show the fifth-order accuracy, great efficiency, high
resolution and robustness of the proposed HWENO scheme.
|
2207.12647 | Yang Liu | Yang Liu, Guanbin Li, Liang Lin | Cross-Modal Causal Relational Reasoning for Event-Level Visual Question
Answering | 17 pages, 9 figures. This work has been accepted by IEEE Transactions
on Pattern Analysis and Machine Intelligence (TPAMI). The datasets, code and
models are available at https://github.com/HCPLab-SYSU/CMCIR | null | null | null | cs.CV cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Existing visual question answering methods often suffer from cross-modal
spurious correlations and oversimplified event-level reasoning processes that
fail to capture event temporality, causality, and dynamics spanning over the
video. In this work, to address the task of event-level visual question
answering, we propose a framework for cross-modal causal relational reasoning.
In particular, a set of causal intervention operations is introduced to
discover the underlying causal structures across visual and linguistic
modalities. Our framework, named Cross-Modal Causal RelatIonal Reasoning
(CMCIR), involves three modules: i) Causality-aware Visual-Linguistic Reasoning
(CVLR) module for collaboratively disentangling the visual and linguistic
spurious correlations via front-door and back-door causal interventions; ii)
Spatial-Temporal Transformer (STT) module for capturing the fine-grained
interactions between visual and linguistic semantics; iii) Visual-Linguistic
Feature Fusion (VLFF) module for learning the global semantic-aware
visual-linguistic representations adaptively. Extensive experiments on four
event-level datasets demonstrate the superiority of our CMCIR in discovering
visual-linguistic causal structures and achieving robust event-level visual
question answering. The datasets, code, and models are available at
https://github.com/HCPLab-SYSU/CMCIR.
| [
{
"version": "v1",
"created": "Tue, 26 Jul 2022 04:25:54 GMT"
},
{
"version": "v2",
"created": "Mon, 15 Aug 2022 08:15:04 GMT"
},
{
"version": "v3",
"created": "Fri, 30 Dec 2022 09:17:12 GMT"
},
{
"version": "v4",
"created": "Sun, 8 Jan 2023 04:09:04 GMT"
},
{
"version": "v5",
"created": "Tue, 18 Apr 2023 03:44:46 GMT"
},
{
"version": "v6",
"created": "Wed, 19 Apr 2023 03:47:14 GMT"
},
{
"version": "v7",
"created": "Tue, 25 Apr 2023 02:47:13 GMT"
},
{
"version": "v8",
"created": "Wed, 7 Jun 2023 07:47:27 GMT"
}
] | 2023-06-08T00:00:00 | [
[
"Liu",
"Yang",
""
],
[
"Li",
"Guanbin",
""
],
[
"Lin",
"Liang",
""
]
] | TITLE: Cross-Modal Causal Relational Reasoning for Event-Level Visual Question
Answering
ABSTRACT: Existing visual question answering methods often suffer from cross-modal
spurious correlations and oversimplified event-level reasoning processes that
fail to capture event temporality, causality, and dynamics spanning over the
video. In this work, to address the task of event-level visual question
answering, we propose a framework for cross-modal causal relational reasoning.
In particular, a set of causal intervention operations is introduced to
discover the underlying causal structures across visual and linguistic
modalities. Our framework, named Cross-Modal Causal RelatIonal Reasoning
(CMCIR), involves three modules: i) Causality-aware Visual-Linguistic Reasoning
(CVLR) module for collaboratively disentangling the visual and linguistic
spurious correlations via front-door and back-door causal interventions; ii)
Spatial-Temporal Transformer (STT) module for capturing the fine-grained
interactions between visual and linguistic semantics; iii) Visual-Linguistic
Feature Fusion (VLFF) module for learning the global semantic-aware
visual-linguistic representations adaptively. Extensive experiments on four
event-level datasets demonstrate the superiority of our CMCIR in discovering
visual-linguistic causal structures and achieving robust event-level visual
question answering. The datasets, code, and models are available at
https://github.com/HCPLab-SYSU/CMCIR.
|
2209.13883 | Mu Yuan | Mu Yuan, Lan Zhang, Zimu Zheng, Yi-Nan Zhang, Xiang-Yang Li | MLink: Linking Black-Box Models from Multiple Domains for Collaborative
Inference | Transactions on Pattern Analysis and Machine Intelligence (TPAMI)
2023 | null | null | null | cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The cost efficiency of model inference is critical to real-world machine
learning (ML) applications, especially for delay-sensitive tasks and
resource-limited devices. A typical dilemma is: in order to provide complex
intelligent services (e.g. smart city), we need inference results of multiple
ML models, but the cost budget (e.g. GPU memory) is not enough to run all of
them. In this work, we study underlying relationships among black-box ML models
and propose a novel learning task: model linking, which aims to bridge the
knowledge of different black-box models by learning mappings (dubbed model
links) between their output spaces. We propose the design of model links which
supports linking heterogeneous black-box ML models. Also, in order to address
the distribution discrepancy challenge, we present adaptation and aggregation
methods of model links. Based on our proposed model links, we developed a
scheduling algorithm, named MLink. Through collaborative multi-model inference
enabled by model links, MLink can improve the accuracy of obtained inference
results under the cost budget. We evaluated MLink on a multi-modal dataset with
seven different ML models and two real-world video analytics systems with six
ML models and 3,264 hours of video. Experimental results show that our proposed
model links can be effectively built among various black-box models. Under the
budget of GPU memory, MLink can save 66.7% inference computations while
preserving 94% inference accuracy, which outperforms multi-task learning, deep
reinforcement learning-based scheduler and frame filtering baselines.
| [
{
"version": "v1",
"created": "Wed, 28 Sep 2022 07:29:47 GMT"
},
{
"version": "v2",
"created": "Tue, 6 Jun 2023 02:14:07 GMT"
},
{
"version": "v3",
"created": "Wed, 7 Jun 2023 03:15:06 GMT"
}
] | 2023-06-08T00:00:00 | [
[
"Yuan",
"Mu",
""
],
[
"Zhang",
"Lan",
""
],
[
"Zheng",
"Zimu",
""
],
[
"Zhang",
"Yi-Nan",
""
],
[
"Li",
"Xiang-Yang",
""
]
] | TITLE: MLink: Linking Black-Box Models from Multiple Domains for Collaborative
Inference
ABSTRACT: The cost efficiency of model inference is critical to real-world machine
learning (ML) applications, especially for delay-sensitive tasks and
resource-limited devices. A typical dilemma is: in order to provide complex
intelligent services (e.g. smart city), we need inference results of multiple
ML models, but the cost budget (e.g. GPU memory) is not enough to run all of
them. In this work, we study underlying relationships among black-box ML models
and propose a novel learning task: model linking, which aims to bridge the
knowledge of different black-box models by learning mappings (dubbed model
links) between their output spaces. We propose the design of model links which
supports linking heterogeneous black-box ML models. Also, in order to address
the distribution discrepancy challenge, we present adaptation and aggregation
methods of model links. Based on our proposed model links, we developed a
scheduling algorithm, named MLink. Through collaborative multi-model inference
enabled by model links, MLink can improve the accuracy of obtained inference
results under the cost budget. We evaluated MLink on a multi-modal dataset with
seven different ML models and two real-world video analytics systems with six
ML models and 3,264 hours of video. Experimental results show that our proposed
model links can be effectively built among various black-box models. Under the
budget of GPU memory, MLink can save 66.7% inference computations while
preserving 94% inference accuracy, which outperforms multi-task learning, deep
reinforcement learning-based scheduler and frame filtering baselines.
|
2210.08374 | Soufian El Yadmani | Soufian El Yadmani, Robin The, Olga Gadyatskaya | Beyond the Surface: Investigating Malicious CVE Proof of Concept
Exploits on GitHub | null | null | null | null | cs.CR | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Exploit proof-of-concepts (PoCs) for known vulnerabilities are widely shared
in the security community. They help security analysts to learn from each other
and they facilitate security assessments and red teaming tasks. In the recent
years, PoCs have been widely distributed, e.g., via dedicated websites and
platforms, and public code repositories such as GitHub. However, there is no
guarantee that PoCs in public code repositories come from trustworthy sources
or even that they do what they are supposed to do.
In this work we investigate GitHub-hosted PoCs for known vulnerabilities
discovered in 2017--2021. We discovered that not all PoCs are trustworthy. Some
proof-of-concepts are malicious, e.g., they attempt to exfiltrate data from the
system they are being run on, or they try to install malware on this system,
and in some cases they have hard-coded reverse shell listener.
To measure the prevalence of this threat, we propose an approach to detecting
malicious PoCs. Our approach relies on the maliciousness symptoms we have
observed in our PoC dataset: calls to malicious IP addresses, encoded malicious
code, and included Trojanized binaries. With this approach, we have discovered
899 malicious repositories out of 47,285 repositories that have been downloaded
and checked (i.e., 1.9% of the studied repositories have indicators of
malicious intent). This figure shows a worrying prevalence of dangerous
malicious PoCs among the exploit code distributed on GitHub.
| [
{
"version": "v1",
"created": "Sat, 15 Oct 2022 20:51:24 GMT"
},
{
"version": "v2",
"created": "Wed, 7 Jun 2023 12:02:10 GMT"
}
] | 2023-06-08T00:00:00 | [
[
"Yadmani",
"Soufian El",
""
],
[
"The",
"Robin",
""
],
[
"Gadyatskaya",
"Olga",
""
]
] | TITLE: Beyond the Surface: Investigating Malicious CVE Proof of Concept
Exploits on GitHub
ABSTRACT: Exploit proof-of-concepts (PoCs) for known vulnerabilities are widely shared
in the security community. They help security analysts to learn from each other
and they facilitate security assessments and red teaming tasks. In the recent
years, PoCs have been widely distributed, e.g., via dedicated websites and
platforms, and public code repositories such as GitHub. However, there is no
guarantee that PoCs in public code repositories come from trustworthy sources
or even that they do what they are supposed to do.
In this work we investigate GitHub-hosted PoCs for known vulnerabilities
discovered in 2017--2021. We discovered that not all PoCs are trustworthy. Some
proof-of-concepts are malicious, e.g., they attempt to exfiltrate data from the
system they are being run on, or they try to install malware on this system,
and in some cases they have hard-coded reverse shell listener.
To measure the prevalence of this threat, we propose an approach to detecting
malicious PoCs. Our approach relies on the maliciousness symptoms we have
observed in our PoC dataset: calls to malicious IP addresses, encoded malicious
code, and included Trojanized binaries. With this approach, we have discovered
899 malicious repositories out of 47,285 repositories that have been downloaded
and checked (i.e., 1.9% of the studied repositories have indicators of
malicious intent). This figure shows a worrying prevalence of dangerous
malicious PoCs among the exploit code distributed on GitHub.
|
2210.09440 | Omid Rohanian | Omid Rohanian, Hannah Jauncey, Mohammadmahdi Nouriborji, Vinod Kumar
Chauhan, Bronner P. Gon\c{c}alves, Christiana Kartsonaki, ISARIC Clinical
Characterisation Group, Laura Merson, David Clifton | Using Bottleneck Adapters to Identify Cancer in Clinical Notes under
Low-Resource Constraints | null | null | null | null | cs.CL cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Processing information locked within clinical health records is a challenging
task that remains an active area of research in biomedical NLP. In this work,
we evaluate a broad set of machine learning techniques ranging from simple RNNs
to specialised transformers such as BioBERT on a dataset containing clinical
notes along with a set of annotations indicating whether a sample is
cancer-related or not.
Furthermore, we specifically employ efficient fine-tuning methods from NLP,
namely, bottleneck adapters and prompt tuning, to adapt the models to our
specialised task. Our evaluations suggest that fine-tuning a frozen BERT model
pre-trained on natural language and with bottleneck adapters outperforms all
other strategies, including full fine-tuning of the specialised BioBERT model.
Based on our findings, we suggest that using bottleneck adapters in
low-resource situations with limited access to labelled data or processing
capacity could be a viable strategy in biomedical text mining. The code used in
the experiments are going to be made available at
https://github.com/omidrohanian/bottleneck-adapters.
| [
{
"version": "v1",
"created": "Mon, 17 Oct 2022 21:22:23 GMT"
},
{
"version": "v2",
"created": "Wed, 7 Jun 2023 10:18:44 GMT"
}
] | 2023-06-08T00:00:00 | [
[
"Rohanian",
"Omid",
""
],
[
"Jauncey",
"Hannah",
""
],
[
"Nouriborji",
"Mohammadmahdi",
""
],
[
"Chauhan",
"Vinod Kumar",
""
],
[
"Gonçalves",
"Bronner P.",
""
],
[
"Kartsonaki",
"Christiana",
""
],
[
"Group",
"ISARIC Clinical Characterisation",
""
],
[
"Merson",
"Laura",
""
],
[
"Clifton",
"David",
""
]
] | TITLE: Using Bottleneck Adapters to Identify Cancer in Clinical Notes under
Low-Resource Constraints
ABSTRACT: Processing information locked within clinical health records is a challenging
task that remains an active area of research in biomedical NLP. In this work,
we evaluate a broad set of machine learning techniques ranging from simple RNNs
to specialised transformers such as BioBERT on a dataset containing clinical
notes along with a set of annotations indicating whether a sample is
cancer-related or not.
Furthermore, we specifically employ efficient fine-tuning methods from NLP,
namely, bottleneck adapters and prompt tuning, to adapt the models to our
specialised task. Our evaluations suggest that fine-tuning a frozen BERT model
pre-trained on natural language and with bottleneck adapters outperforms all
other strategies, including full fine-tuning of the specialised BioBERT model.
Based on our findings, we suggest that using bottleneck adapters in
low-resource situations with limited access to labelled data or processing
capacity could be a viable strategy in biomedical text mining. The code used in
the experiments are going to be made available at
https://github.com/omidrohanian/bottleneck-adapters.
|
2211.02918 | Xiao Chi | Xiao Chi | A Filtering-based General Approach to Learning Rational Constraints of
Epistemic Graphs | 18 pages, 6 figures, submitted to CLAR 2023 | null | null | null | cs.AI cs.LG | http://creativecommons.org/licenses/by/4.0/ | Epistemic graphs are a generalization of the epistemic approach to
probabilistic argumentation. Hunter proposed a 2-way generalization framework
to learn epistemic constraints from crowd-sourcing data. However, the learnt
epistemic constraints only reflect users' beliefs from data, without
considering the rationality encoded in epistemic graphs. Meanwhile, the current
framework can only generate epistemic constraints that reflect whether an agent
believes an argument, but not the degree to which it believes in it. The major
challenge to achieving this effect is that the computational complexity will
increase sharply when expanding the variety of constraints, which may lead to
unacceptable time performance. To address these problems, we propose a
filtering-based approach using a multiple-way generalization step to generate a
set of rational rules which are consistent with their epistemic graphs from a
dataset. This approach is able to learn a wider variety of rational rules that
reflect information in both the domain model and the user model. Moreover, to
improve computational efficiency, we introduce a new function to exclude
meaningless rules. The empirical results show that our approach significantly
outperforms the existing framework when expanding the variety of rules.
| [
{
"version": "v1",
"created": "Sat, 5 Nov 2022 14:24:46 GMT"
},
{
"version": "v2",
"created": "Wed, 7 Jun 2023 15:13:33 GMT"
}
] | 2023-06-08T00:00:00 | [
[
"Chi",
"Xiao",
""
]
] | TITLE: A Filtering-based General Approach to Learning Rational Constraints of
Epistemic Graphs
ABSTRACT: Epistemic graphs are a generalization of the epistemic approach to
probabilistic argumentation. Hunter proposed a 2-way generalization framework
to learn epistemic constraints from crowd-sourcing data. However, the learnt
epistemic constraints only reflect users' beliefs from data, without
considering the rationality encoded in epistemic graphs. Meanwhile, the current
framework can only generate epistemic constraints that reflect whether an agent
believes an argument, but not the degree to which it believes in it. The major
challenge to achieving this effect is that the computational complexity will
increase sharply when expanding the variety of constraints, which may lead to
unacceptable time performance. To address these problems, we propose a
filtering-based approach using a multiple-way generalization step to generate a
set of rational rules which are consistent with their epistemic graphs from a
dataset. This approach is able to learn a wider variety of rational rules that
reflect information in both the domain model and the user model. Moreover, to
improve computational efficiency, we introduce a new function to exclude
meaningless rules. The empirical results show that our approach significantly
outperforms the existing framework when expanding the variety of rules.
|
2211.05041 | Yusen Zhang | Yusen Zhang, Yang Liu, Ziyi Yang, Yuwei Fang, Yulong Chen, Dragomir
Radev, Chenguang Zhu, Michael Zeng, Rui Zhang | MACSum: Controllable Summarization with Mixed Attributes | TACL 2023 | null | null | null | cs.CL | http://creativecommons.org/licenses/by/4.0/ | Controllable summarization allows users to generate customized summaries with
specified attributes. However, due to the lack of designated annotations of
controlled summaries, existing works have to craft pseudo datasets by adapting
generic summarization benchmarks. Furthermore, most research focuses on
controlling single attributes individually (e.g., a short summary or a highly
abstractive summary) rather than controlling a mix of attributes together
(e.g., a short and highly abstractive summary). In this paper, we propose
MACSum, the first human-annotated summarization dataset for controlling mixed
attributes. It contains source texts from two domains, news articles and
dialogues, with human-annotated summaries controlled by five designed
attributes (Length, Extractiveness, Specificity, Topic, and Speaker). We
propose two simple and effective parameter-efficient approaches for the new
task of mixed controllable summarization based on hard prompt tuning and soft
prefix tuning. Results and analysis demonstrate that hard prompt models yield
the best performance on all metrics and human evaluations. However,
mixed-attribute control is still challenging for summarization tasks. Our
dataset and code are available at https://github.com/psunlpgroup/MACSum.
| [
{
"version": "v1",
"created": "Wed, 9 Nov 2022 17:17:37 GMT"
},
{
"version": "v2",
"created": "Wed, 7 Jun 2023 02:02:51 GMT"
}
] | 2023-06-08T00:00:00 | [
[
"Zhang",
"Yusen",
""
],
[
"Liu",
"Yang",
""
],
[
"Yang",
"Ziyi",
""
],
[
"Fang",
"Yuwei",
""
],
[
"Chen",
"Yulong",
""
],
[
"Radev",
"Dragomir",
""
],
[
"Zhu",
"Chenguang",
""
],
[
"Zeng",
"Michael",
""
],
[
"Zhang",
"Rui",
""
]
] | TITLE: MACSum: Controllable Summarization with Mixed Attributes
ABSTRACT: Controllable summarization allows users to generate customized summaries with
specified attributes. However, due to the lack of designated annotations of
controlled summaries, existing works have to craft pseudo datasets by adapting
generic summarization benchmarks. Furthermore, most research focuses on
controlling single attributes individually (e.g., a short summary or a highly
abstractive summary) rather than controlling a mix of attributes together
(e.g., a short and highly abstractive summary). In this paper, we propose
MACSum, the first human-annotated summarization dataset for controlling mixed
attributes. It contains source texts from two domains, news articles and
dialogues, with human-annotated summaries controlled by five designed
attributes (Length, Extractiveness, Specificity, Topic, and Speaker). We
propose two simple and effective parameter-efficient approaches for the new
task of mixed controllable summarization based on hard prompt tuning and soft
prefix tuning. Results and analysis demonstrate that hard prompt models yield
the best performance on all metrics and human evaluations. However,
mixed-attribute control is still challenging for summarization tasks. Our
dataset and code are available at https://github.com/psunlpgroup/MACSum.
|
2211.14666 | S\'ebastien Lachapelle | S\'ebastien Lachapelle, Tristan Deleu, Divyat Mahajan, Ioannis
Mitliagkas, Yoshua Bengio, Simon Lacoste-Julien, Quentin Bertrand | Synergies between Disentanglement and Sparsity: Generalization and
Identifiability in Multi-Task Learning | Appears in: Fortieth International Conference on Machine Learning
(ICML 2023). 36 pages | null | null | null | cs.LG stat.ML | http://creativecommons.org/licenses/by/4.0/ | Although disentangled representations are often said to be beneficial for
downstream tasks, current empirical and theoretical understanding is limited.
In this work, we provide evidence that disentangled representations coupled
with sparse base-predictors improve generalization. In the context of
multi-task learning, we prove a new identifiability result that provides
conditions under which maximally sparse base-predictors yield disentangled
representations. Motivated by this theoretical result, we propose a practical
approach to learn disentangled representations based on a sparsity-promoting
bi-level optimization problem. Finally, we explore a meta-learning version of
this algorithm based on group Lasso multiclass SVM base-predictors, for which
we derive a tractable dual formulation. It obtains competitive results on
standard few-shot classification benchmarks, while each task is using only a
fraction of the learned representations.
| [
{
"version": "v1",
"created": "Sat, 26 Nov 2022 21:02:09 GMT"
},
{
"version": "v2",
"created": "Tue, 6 Jun 2023 18:02:14 GMT"
}
] | 2023-06-08T00:00:00 | [
[
"Lachapelle",
"Sébastien",
""
],
[
"Deleu",
"Tristan",
""
],
[
"Mahajan",
"Divyat",
""
],
[
"Mitliagkas",
"Ioannis",
""
],
[
"Bengio",
"Yoshua",
""
],
[
"Lacoste-Julien",
"Simon",
""
],
[
"Bertrand",
"Quentin",
""
]
] | TITLE: Synergies between Disentanglement and Sparsity: Generalization and
Identifiability in Multi-Task Learning
ABSTRACT: Although disentangled representations are often said to be beneficial for
downstream tasks, current empirical and theoretical understanding is limited.
In this work, we provide evidence that disentangled representations coupled
with sparse base-predictors improve generalization. In the context of
multi-task learning, we prove a new identifiability result that provides
conditions under which maximally sparse base-predictors yield disentangled
representations. Motivated by this theoretical result, we propose a practical
approach to learn disentangled representations based on a sparsity-promoting
bi-level optimization problem. Finally, we explore a meta-learning version of
this algorithm based on group Lasso multiclass SVM base-predictors, for which
we derive a tractable dual formulation. It obtains competitive results on
standard few-shot classification benchmarks, while each task is using only a
fraction of the learned representations.
|
2211.16582 | Vladimir Kulikov | Vladimir Kulikov, Shahar Yadin, Matan Kleiner, Tomer Michaeli | SinDDM: A Single Image Denoising Diffusion Model | Updated for ICML 2023 and added the Appendix. Note that the images
are lightly compressed. Visit our project page for uncompressed results:
https://matankleiner.github.io/sinddm/ | null | null | null | cs.CV cs.LG eess.IV | http://creativecommons.org/licenses/by/4.0/ | Denoising diffusion models (DDMs) have led to staggering performance leaps in
image generation, editing and restoration. However, existing DDMs use very
large datasets for training. Here, we introduce a framework for training a DDM
on a single image. Our method, which we coin SinDDM, learns the internal
statistics of the training image by using a multi-scale diffusion process. To
drive the reverse diffusion process, we use a fully-convolutional light-weight
denoiser, which is conditioned on both the noise level and the scale. This
architecture allows generating samples of arbitrary dimensions, in a
coarse-to-fine manner. As we illustrate, SinDDM generates diverse high-quality
samples, and is applicable in a wide array of tasks, including style transfer
and harmonization. Furthermore, it can be easily guided by external
supervision. Particularly, we demonstrate text-guided generation from a single
image using a pre-trained CLIP model.
| [
{
"version": "v1",
"created": "Tue, 29 Nov 2022 20:44:25 GMT"
},
{
"version": "v2",
"created": "Wed, 28 Dec 2022 18:36:51 GMT"
},
{
"version": "v3",
"created": "Tue, 6 Jun 2023 20:42:41 GMT"
}
] | 2023-06-08T00:00:00 | [
[
"Kulikov",
"Vladimir",
""
],
[
"Yadin",
"Shahar",
""
],
[
"Kleiner",
"Matan",
""
],
[
"Michaeli",
"Tomer",
""
]
] | TITLE: SinDDM: A Single Image Denoising Diffusion Model
ABSTRACT: Denoising diffusion models (DDMs) have led to staggering performance leaps in
image generation, editing and restoration. However, existing DDMs use very
large datasets for training. Here, we introduce a framework for training a DDM
on a single image. Our method, which we coin SinDDM, learns the internal
statistics of the training image by using a multi-scale diffusion process. To
drive the reverse diffusion process, we use a fully-convolutional light-weight
denoiser, which is conditioned on both the noise level and the scale. This
architecture allows generating samples of arbitrary dimensions, in a
coarse-to-fine manner. As we illustrate, SinDDM generates diverse high-quality
samples, and is applicable in a wide array of tasks, including style transfer
and harmonization. Furthermore, it can be easily guided by external
supervision. Particularly, we demonstrate text-guided generation from a single
image using a pre-trained CLIP model.
|
2212.03752 | Qingyan Bai | Qingyan Bai, Ceyuan Yang, Yinghao Xu, Xihui Liu, Yujiu Yang, Yujun
Shen | GLeaD: Improving GANs with A Generator-Leading Task | CVPR2023. Project page: https://ezioby.github.io/glead/ Code:
https://github.com/EzioBy/glead/ | null | null | null | cs.CV eess.IV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Generative adversarial network (GAN) is formulated as a two-player game
between a generator (G) and a discriminator (D), where D is asked to
differentiate whether an image comes from real data or is produced by G. Under
such a formulation, D plays as the rule maker and hence tends to dominate the
competition. Towards a fairer game in GANs, we propose a new paradigm for
adversarial training, which makes G assign a task to D as well. Specifically,
given an image, we expect D to extract representative features that can be
adequately decoded by G to reconstruct the input. That way, instead of learning
freely, D is urged to align with the view of G for domain classification.
Experimental results on various datasets demonstrate the substantial
superiority of our approach over the baselines. For instance, we improve the
FID of StyleGAN2 from 4.30 to 2.55 on LSUN Bedroom and from 4.04 to 2.82 on
LSUN Church. We believe that the pioneering attempt present in this work could
inspire the community with better designed generator-leading tasks for GAN
improvement.
| [
{
"version": "v1",
"created": "Wed, 7 Dec 2022 16:25:19 GMT"
},
{
"version": "v2",
"created": "Wed, 7 Jun 2023 03:34:34 GMT"
}
] | 2023-06-08T00:00:00 | [
[
"Bai",
"Qingyan",
""
],
[
"Yang",
"Ceyuan",
""
],
[
"Xu",
"Yinghao",
""
],
[
"Liu",
"Xihui",
""
],
[
"Yang",
"Yujiu",
""
],
[
"Shen",
"Yujun",
""
]
] | TITLE: GLeaD: Improving GANs with A Generator-Leading Task
ABSTRACT: Generative adversarial network (GAN) is formulated as a two-player game
between a generator (G) and a discriminator (D), where D is asked to
differentiate whether an image comes from real data or is produced by G. Under
such a formulation, D plays as the rule maker and hence tends to dominate the
competition. Towards a fairer game in GANs, we propose a new paradigm for
adversarial training, which makes G assign a task to D as well. Specifically,
given an image, we expect D to extract representative features that can be
adequately decoded by G to reconstruct the input. That way, instead of learning
freely, D is urged to align with the view of G for domain classification.
Experimental results on various datasets demonstrate the substantial
superiority of our approach over the baselines. For instance, we improve the
FID of StyleGAN2 from 4.30 to 2.55 on LSUN Bedroom and from 4.04 to 2.82 on
LSUN Church. We believe that the pioneering attempt present in this work could
inspire the community with better designed generator-leading tasks for GAN
improvement.
|
2212.08427 | Javier Cabrera Arteaga | Javier Cabrera-Arteaga, Martin Monperrus, Tim Toady and Benoit Baudry | WebAssembly Diversification for Malware Evasion | null | Computers & Security, 2023 | 10.1016/j.cose.2023.103296 | null | cs.SE | http://creativecommons.org/licenses/by/4.0/ | WebAssembly has become a crucial part of the modern web, offering a faster
alternative to JavaScript in browsers. While boosting rich applications in
browser, this technology is also very efficient to develop cryptojacking
malware. This has triggered the development of several methods to detect
cryptojacking malware. However, these defenses have not considered the
possibility of attackers using evasion techniques. This paper explores how
automatic binary diversification can support the evasion of WebAssembly
cryptojacking detectors. We experiment with a dataset of 33 WebAssembly
cryptojacking binaries and evaluate our evasion technique against two malware
detectors: VirusTotal, a general-purpose detector, and MINOS, a
WebAssembly-specific detector. Our results demonstrate that our technique can
automatically generate variants of WebAssembly cryptojacking that evade the
detectors in 90% of cases for VirusTotal and 100% for MINOS. Our results
emphasize the importance of meta-antiviruses and diverse detection techniques,
and provide new insights into which WebAssembly code transformations are best
suited for malware evasion. We also show that the variants introduce limited
performance overhead, making binary diversification an effective technique for
evasion.
| [
{
"version": "v1",
"created": "Fri, 16 Dec 2022 12:02:26 GMT"
},
{
"version": "v2",
"created": "Thu, 27 Apr 2023 05:53:26 GMT"
}
] | 2023-06-08T00:00:00 | [
[
"Cabrera-Arteaga",
"Javier",
""
],
[
"Monperrus",
"Martin",
""
],
[
"Toady",
"Tim",
""
],
[
"Baudry",
"Benoit",
""
]
] | TITLE: WebAssembly Diversification for Malware Evasion
ABSTRACT: WebAssembly has become a crucial part of the modern web, offering a faster
alternative to JavaScript in browsers. While boosting rich applications in
browser, this technology is also very efficient to develop cryptojacking
malware. This has triggered the development of several methods to detect
cryptojacking malware. However, these defenses have not considered the
possibility of attackers using evasion techniques. This paper explores how
automatic binary diversification can support the evasion of WebAssembly
cryptojacking detectors. We experiment with a dataset of 33 WebAssembly
cryptojacking binaries and evaluate our evasion technique against two malware
detectors: VirusTotal, a general-purpose detector, and MINOS, a
WebAssembly-specific detector. Our results demonstrate that our technique can
automatically generate variants of WebAssembly cryptojacking that evade the
detectors in 90% of cases for VirusTotal and 100% for MINOS. Our results
emphasize the importance of meta-antiviruses and diverse detection techniques,
and provide new insights into which WebAssembly code transformations are best
suited for malware evasion. We also show that the variants introduce limited
performance overhead, making binary diversification an effective technique for
evasion.
|
2212.09567 | Yushi Bai | Yushi Bai, Xin Lv, Juanzi Li, Lei Hou | Answering Complex Logical Queries on Knowledge Graphs via Query
Computation Tree Optimization | To appear in ICML 2023 | null | null | null | cs.LG cs.AI cs.DB cs.SI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Answering complex logical queries on incomplete knowledge graphs is a
challenging task, and has been widely studied. Embedding-based methods require
training on complex queries, and cannot generalize well to out-of-distribution
query structures. Recent work frames this task as an end-to-end optimization
problem, and it only requires a pretrained link predictor. However, due to the
exponentially large combinatorial search space, the optimal solution can only
be approximated, limiting the final accuracy. In this work, we propose QTO
(Query Computation Tree Optimization) that can efficiently find the exact
optimal solution. QTO finds the optimal solution by a forward-backward
propagation on the tree-like computation graph, i.e., query computation tree.
In particular, QTO utilizes the independence encoded in the query computation
tree to reduce the search space, where only local computations are involved
during the optimization procedure. Experiments on 3 datasets show that QTO
obtains state-of-the-art performance on complex query answering, outperforming
previous best results by an average of 22%. Moreover, QTO can interpret the
intermediate solutions for each of the one-hop atoms in the query with over 90%
accuracy. The code of our paper is at https://github.com/bys0318/QTO.
| [
{
"version": "v1",
"created": "Mon, 19 Dec 2022 15:59:00 GMT"
},
{
"version": "v2",
"created": "Wed, 21 Dec 2022 06:57:32 GMT"
},
{
"version": "v3",
"created": "Wed, 7 Jun 2023 06:57:09 GMT"
}
] | 2023-06-08T00:00:00 | [
[
"Bai",
"Yushi",
""
],
[
"Lv",
"Xin",
""
],
[
"Li",
"Juanzi",
""
],
[
"Hou",
"Lei",
""
]
] | TITLE: Answering Complex Logical Queries on Knowledge Graphs via Query
Computation Tree Optimization
ABSTRACT: Answering complex logical queries on incomplete knowledge graphs is a
challenging task, and has been widely studied. Embedding-based methods require
training on complex queries, and cannot generalize well to out-of-distribution
query structures. Recent work frames this task as an end-to-end optimization
problem, and it only requires a pretrained link predictor. However, due to the
exponentially large combinatorial search space, the optimal solution can only
be approximated, limiting the final accuracy. In this work, we propose QTO
(Query Computation Tree Optimization) that can efficiently find the exact
optimal solution. QTO finds the optimal solution by a forward-backward
propagation on the tree-like computation graph, i.e., query computation tree.
In particular, QTO utilizes the independence encoded in the query computation
tree to reduce the search space, where only local computations are involved
during the optimization procedure. Experiments on 3 datasets show that QTO
obtains state-of-the-art performance on complex query answering, outperforming
previous best results by an average of 22%. Moreover, QTO can interpret the
intermediate solutions for each of the one-hop atoms in the query with over 90%
accuracy. The code of our paper is at https://github.com/bys0318/QTO.
|
2212.09737 | Jinpeng Wang | Alex Jinpeng Wang, Pan Zhou, Mike Zheng Shou, Shuicheng Yan | Position-guided Text Prompt for Vision-Language Pre-training | Camera-ready version, code is in https://github.com/sail-sg/ptp | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Vision-Language Pre-Training (VLP) has shown promising capabilities to align
image and text pairs, facilitating a broad variety of cross-modal learning
tasks. However, we observe that VLP models often lack the visual
grounding/localization capability which is critical for many downstream tasks
such as visual reasoning. In this work, we propose a novel Position-guided Text
Prompt (PTP) paradigm to enhance the visual grounding ability of cross-modal
models trained with VLP. Specifically, in the VLP phase, PTP divides the image
into $N\times N$ blocks, and identifies the objects in each block through the
widely used object detector in VLP. It then reformulates the visual grounding
task into a fill-in-the-blank problem given a PTP by encouraging the model to
predict the objects in the given blocks or regress the blocks of a given
object, e.g. filling `P" or ``O" in aPTP ``The block P has a O". This mechanism
improves the visual grounding capability of VLP models and thus helps them
better handle various downstream tasks. By introducing PTP into several
state-of-the-art VLP frameworks, we observe consistently significant
improvements across representative cross-modal learning model architectures and
several benchmarks, e.g. zero-shot Flickr30K Retrieval (+4.8 in average
recall@1) for ViLT \cite{vilt} baseline, and COCO Captioning (+5.3 in CIDEr)
for SOTA BLIP \cite{blip} baseline. Moreover, PTP achieves comparable results
with object-detector based methods, and much faster inference speed since PTP
discards its object detector for inference while the later cannot. Our code and
pre-trained weight will be released at \url{https://github.com/sail-sg/ptp}.
| [
{
"version": "v1",
"created": "Mon, 19 Dec 2022 18:55:43 GMT"
},
{
"version": "v2",
"created": "Wed, 7 Jun 2023 06:28:18 GMT"
}
] | 2023-06-08T00:00:00 | [
[
"Wang",
"Alex Jinpeng",
""
],
[
"Zhou",
"Pan",
""
],
[
"Shou",
"Mike Zheng",
""
],
[
"Yan",
"Shuicheng",
""
]
] | TITLE: Position-guided Text Prompt for Vision-Language Pre-training
ABSTRACT: Vision-Language Pre-Training (VLP) has shown promising capabilities to align
image and text pairs, facilitating a broad variety of cross-modal learning
tasks. However, we observe that VLP models often lack the visual
grounding/localization capability which is critical for many downstream tasks
such as visual reasoning. In this work, we propose a novel Position-guided Text
Prompt (PTP) paradigm to enhance the visual grounding ability of cross-modal
models trained with VLP. Specifically, in the VLP phase, PTP divides the image
into $N\times N$ blocks, and identifies the objects in each block through the
widely used object detector in VLP. It then reformulates the visual grounding
task into a fill-in-the-blank problem given a PTP by encouraging the model to
predict the objects in the given blocks or regress the blocks of a given
object, e.g. filling `P" or ``O" in aPTP ``The block P has a O". This mechanism
improves the visual grounding capability of VLP models and thus helps them
better handle various downstream tasks. By introducing PTP into several
state-of-the-art VLP frameworks, we observe consistently significant
improvements across representative cross-modal learning model architectures and
several benchmarks, e.g. zero-shot Flickr30K Retrieval (+4.8 in average
recall@1) for ViLT \cite{vilt} baseline, and COCO Captioning (+5.3 in CIDEr)
for SOTA BLIP \cite{blip} baseline. Moreover, PTP achieves comparable results
with object-detector based methods, and much faster inference speed since PTP
discards its object detector for inference while the later cannot. Our code and
pre-trained weight will be released at \url{https://github.com/sail-sg/ptp}.
|
2301.07053 | Jo\"el L. Lavanchy | Jo\"el L. Lavanchy, Armine Vardazaryan, Pietro Mascagni, AI4SafeChole
Consortium, Didier Mutter, Nicolas Padoy | Preserving Privacy in Surgical Video Analysis Using Artificial
Intelligence: A Deep Learning Classifier to Identify Out-of-Body Scenes in
Endoscopic Videos | Jo\"el L. Lavanchy and Armine Vardazaryan contributed equally and
share first co-authorship | Scientific Reports 13, 9235 (2023) | 10.1038/s41598-023-36453-1 | null | cs.CV | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Objective: To develop and validate a deep learning model for the
identification of out-of-body images in endoscopic videos. Background: Surgical
video analysis facilitates education and research. However, video recordings of
endoscopic surgeries can contain privacy-sensitive information, especially if
out-of-body scenes are recorded. Therefore, identification of out-of-body
scenes in endoscopic videos is of major importance to preserve the privacy of
patients and operating room staff. Methods: A deep learning model was trained
and evaluated on an internal dataset of 12 different types of laparoscopic and
robotic surgeries. External validation was performed on two independent
multicentric test datasets of laparoscopic gastric bypass and cholecystectomy
surgeries. All images extracted from the video datasets were annotated as
inside or out-of-body. Model performance was evaluated compared to human ground
truth annotations measuring the receiver operating characteristic area under
the curve (ROC AUC). Results: The internal dataset consisting of 356,267 images
from 48 videos and the two multicentric test datasets consisting of 54,385 and
58,349 images from 10 and 20 videos, respectively, were annotated. Compared to
ground truth annotations, the model identified out-of-body images with 99.97%
ROC AUC on the internal test dataset. Mean $\pm$ standard deviation ROC AUC on
the multicentric gastric bypass dataset was 99.94$\pm$0.07% and 99.71$\pm$0.40%
on the multicentric cholecystectomy dataset, respectively. Conclusion: The
proposed deep learning model can reliably identify out-of-body images in
endoscopic videos. The trained model is publicly shared. This facilitates
privacy preservation in surgical video analysis.
| [
{
"version": "v1",
"created": "Tue, 17 Jan 2023 18:09:44 GMT"
},
{
"version": "v2",
"created": "Wed, 7 Jun 2023 11:02:35 GMT"
}
] | 2023-06-08T00:00:00 | [
[
"Lavanchy",
"Joël L.",
""
],
[
"Vardazaryan",
"Armine",
""
],
[
"Mascagni",
"Pietro",
""
],
[
"Consortium",
"AI4SafeChole",
""
],
[
"Mutter",
"Didier",
""
],
[
"Padoy",
"Nicolas",
""
]
] | TITLE: Preserving Privacy in Surgical Video Analysis Using Artificial
Intelligence: A Deep Learning Classifier to Identify Out-of-Body Scenes in
Endoscopic Videos
ABSTRACT: Objective: To develop and validate a deep learning model for the
identification of out-of-body images in endoscopic videos. Background: Surgical
video analysis facilitates education and research. However, video recordings of
endoscopic surgeries can contain privacy-sensitive information, especially if
out-of-body scenes are recorded. Therefore, identification of out-of-body
scenes in endoscopic videos is of major importance to preserve the privacy of
patients and operating room staff. Methods: A deep learning model was trained
and evaluated on an internal dataset of 12 different types of laparoscopic and
robotic surgeries. External validation was performed on two independent
multicentric test datasets of laparoscopic gastric bypass and cholecystectomy
surgeries. All images extracted from the video datasets were annotated as
inside or out-of-body. Model performance was evaluated compared to human ground
truth annotations measuring the receiver operating characteristic area under
the curve (ROC AUC). Results: The internal dataset consisting of 356,267 images
from 48 videos and the two multicentric test datasets consisting of 54,385 and
58,349 images from 10 and 20 videos, respectively, were annotated. Compared to
ground truth annotations, the model identified out-of-body images with 99.97%
ROC AUC on the internal test dataset. Mean $\pm$ standard deviation ROC AUC on
the multicentric gastric bypass dataset was 99.94$\pm$0.07% and 99.71$\pm$0.40%
on the multicentric cholecystectomy dataset, respectively. Conclusion: The
proposed deep learning model can reliably identify out-of-body images in
endoscopic videos. The trained model is publicly shared. This facilitates
privacy preservation in surgical video analysis.
|
2302.01228 | Rasmus Kj{\ae}r H{\o}ier | Rasmus H{\o}ier, D. Staudt, Christopher Zach | Dual Propagation: Accelerating Contrastive Hebbian Learning with Dyadic
Neurons | Added reflections on biological plausibility and results comparisons
to state-of-the-art versions of equilibrium propagation and difference target
propagation | null | null | null | cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Activity difference based learning algorithms-such as contrastive Hebbian
learning and equilibrium propagation-have been proposed as biologically
plausible alternatives to error back-propagation. However, on traditional
digital chips these algorithms suffer from having to solve a costly inference
problem twice, making these approaches more than two orders of magnitude slower
than back-propagation. In the analog realm equilibrium propagation may be
promising for fast and energy efficient learning, but states still need to be
inferred and stored twice. Inspired by lifted neural networks and compartmental
neuron models we propose a simple energy based compartmental neuron model,
termed dual propagation, in which each neuron is a dyad with two intrinsic
states. At inference time these intrinsic states encode the error/activity
duality through their difference and their mean respectively. The advantage of
this method is that only a single inference phase is needed and that inference
can be solved in layerwise closed-form. Experimentally we show on common
computer vision datasets, including Imagenet32x32, that dual propagation
performs equivalently to back-propagation both in terms of accuracy and
runtime.
| [
{
"version": "v1",
"created": "Thu, 2 Feb 2023 17:07:36 GMT"
},
{
"version": "v2",
"created": "Thu, 23 Feb 2023 13:21:37 GMT"
},
{
"version": "v3",
"created": "Wed, 7 Jun 2023 08:48:04 GMT"
}
] | 2023-06-08T00:00:00 | [
[
"Høier",
"Rasmus",
""
],
[
"Staudt",
"D.",
""
],
[
"Zach",
"Christopher",
""
]
] | TITLE: Dual Propagation: Accelerating Contrastive Hebbian Learning with Dyadic
Neurons
ABSTRACT: Activity difference based learning algorithms-such as contrastive Hebbian
learning and equilibrium propagation-have been proposed as biologically
plausible alternatives to error back-propagation. However, on traditional
digital chips these algorithms suffer from having to solve a costly inference
problem twice, making these approaches more than two orders of magnitude slower
than back-propagation. In the analog realm equilibrium propagation may be
promising for fast and energy efficient learning, but states still need to be
inferred and stored twice. Inspired by lifted neural networks and compartmental
neuron models we propose a simple energy based compartmental neuron model,
termed dual propagation, in which each neuron is a dyad with two intrinsic
states. At inference time these intrinsic states encode the error/activity
duality through their difference and their mean respectively. The advantage of
this method is that only a single inference phase is needed and that inference
can be solved in layerwise closed-form. Experimentally we show on common
computer vision datasets, including Imagenet32x32, that dual propagation
performs equivalently to back-propagation both in terms of accuracy and
runtime.
|
2302.05195 | Thomas Stegm\"uller | Thomas Stegm\"uller, Christian Abbet, Behzad Bozorgtabar, Holly
Clarke, Patrick Petignat, Pierre Vassilakos, and Jean-Philippe Thiran | Self-supervised learning-based cervical cytology for the triage of
HPV-positive women in resource-limited settings and low-data regime | null | null | null | null | eess.IV cs.CV | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Screening Papanicolaou test samples has proven to be highly effective in
reducing cervical cancer-related mortality. However, the lack of trained
cytopathologists hinders its widespread implementation in low-resource
settings. Deep learning-based telecytology diagnosis emerges as an appealing
alternative, but it requires the collection of large annotated training
datasets, which is costly and time-consuming. In this paper, we demonstrate
that the abundance of unlabeled images that can be extracted from Pap smear
test whole slide images presents a fertile ground for self-supervised learning
methods, yielding performance improvements relative to readily available
pre-trained models for various downstream tasks. In particular, we propose
\textbf{C}ervical \textbf{C}ell \textbf{C}opy-\textbf{P}asting
($\texttt{C}^{3}\texttt{P}$) as an effective augmentation method, which enables
knowledge transfer from open-source and labeled single-cell datasets to
unlabeled tiles. Not only does $\texttt{C}^{3}\texttt{P}$ outperforms naive
transfer from single-cell images, but we also demonstrate its advantageous
integration into multiple instance learning methods. Importantly, all our
experiments are conducted on our introduced \textit{in-house} dataset
comprising liquid-based cytology Pap smear images obtained using low-cost
technologies. This aligns with our objective of leveraging deep learning-based
telecytology for diagnosis in low-resource settings.
| [
{
"version": "v1",
"created": "Fri, 10 Feb 2023 11:46:48 GMT"
},
{
"version": "v2",
"created": "Wed, 7 Jun 2023 16:05:44 GMT"
}
] | 2023-06-08T00:00:00 | [
[
"Stegmüller",
"Thomas",
""
],
[
"Abbet",
"Christian",
""
],
[
"Bozorgtabar",
"Behzad",
""
],
[
"Clarke",
"Holly",
""
],
[
"Petignat",
"Patrick",
""
],
[
"Vassilakos",
"Pierre",
""
],
[
"Thiran",
"Jean-Philippe",
""
]
] | TITLE: Self-supervised learning-based cervical cytology for the triage of
HPV-positive women in resource-limited settings and low-data regime
ABSTRACT: Screening Papanicolaou test samples has proven to be highly effective in
reducing cervical cancer-related mortality. However, the lack of trained
cytopathologists hinders its widespread implementation in low-resource
settings. Deep learning-based telecytology diagnosis emerges as an appealing
alternative, but it requires the collection of large annotated training
datasets, which is costly and time-consuming. In this paper, we demonstrate
that the abundance of unlabeled images that can be extracted from Pap smear
test whole slide images presents a fertile ground for self-supervised learning
methods, yielding performance improvements relative to readily available
pre-trained models for various downstream tasks. In particular, we propose
\textbf{C}ervical \textbf{C}ell \textbf{C}opy-\textbf{P}asting
($\texttt{C}^{3}\texttt{P}$) as an effective augmentation method, which enables
knowledge transfer from open-source and labeled single-cell datasets to
unlabeled tiles. Not only does $\texttt{C}^{3}\texttt{P}$ outperforms naive
transfer from single-cell images, but we also demonstrate its advantageous
integration into multiple instance learning methods. Importantly, all our
experiments are conducted on our introduced \textit{in-house} dataset
comprising liquid-based cytology Pap smear images obtained using low-cost
technologies. This aligns with our objective of leveraging deep learning-based
telecytology for diagnosis in low-resource settings.
|
2302.08950 | Vinicius Ribeiro | Vinicius Ribeiro, Yiteng Huang, Yuan Shangguan, Zhaojun Yang, Li Wan,
Ming Sun | Handling the Alignment for Wake Word Detection: A Comparison Between
Alignment-Based, Alignment-Free and Hybrid Approaches | Accepted to Interspeech 2023 | null | null | null | cs.CL cs.SD eess.AS | http://creativecommons.org/licenses/by/4.0/ | Wake word detection exists in most intelligent homes and portable devices. It
offers these devices the ability to "wake up" when summoned at a low cost of
power and computing. This paper focuses on understanding alignment's role in
developing a wake-word system that answers a generic phrase. We discuss three
approaches. The first is alignment-based, where the model is trained with
frame-wise cross-entropy. The second is alignment-free, where the model is
trained with CTC. The third, proposed by us, is a hybrid solution in which the
model is trained with a small set of aligned data and then tuned with a
sizeable unaligned dataset. We compare the three approaches and evaluate the
impact of the different aligned-to-unaligned ratios for hybrid training. Our
results show that the alignment-free system performs better than the
alignment-based for the target operating point, and with a small fraction of
the data (20%), we can train a model that complies with our initial
constraints.
| [
{
"version": "v1",
"created": "Fri, 17 Feb 2023 15:33:47 GMT"
},
{
"version": "v2",
"created": "Mon, 22 May 2023 14:44:21 GMT"
},
{
"version": "v3",
"created": "Wed, 7 Jun 2023 15:04:41 GMT"
}
] | 2023-06-08T00:00:00 | [
[
"Ribeiro",
"Vinicius",
""
],
[
"Huang",
"Yiteng",
""
],
[
"Shangguan",
"Yuan",
""
],
[
"Yang",
"Zhaojun",
""
],
[
"Wan",
"Li",
""
],
[
"Sun",
"Ming",
""
]
] | TITLE: Handling the Alignment for Wake Word Detection: A Comparison Between
Alignment-Based, Alignment-Free and Hybrid Approaches
ABSTRACT: Wake word detection exists in most intelligent homes and portable devices. It
offers these devices the ability to "wake up" when summoned at a low cost of
power and computing. This paper focuses on understanding alignment's role in
developing a wake-word system that answers a generic phrase. We discuss three
approaches. The first is alignment-based, where the model is trained with
frame-wise cross-entropy. The second is alignment-free, where the model is
trained with CTC. The third, proposed by us, is a hybrid solution in which the
model is trained with a small set of aligned data and then tuned with a
sizeable unaligned dataset. We compare the three approaches and evaluate the
impact of the different aligned-to-unaligned ratios for hybrid training. Our
results show that the alignment-free system performs better than the
alignment-based for the target operating point, and with a small fraction of
the data (20%), we can train a model that complies with our initial
constraints.
|
2303.01433 | Aaditya Naik | Aaditya Naik, Yinjun Wu, Mayur Naik, Eric Wong | Do Machine Learning Models Learn Statistical Rules Inferred from Data? | null | null | null | null | cs.LG cs.AI | http://creativecommons.org/licenses/by/4.0/ | Machine learning models can make critical errors that are easily hidden
within vast amounts of data. Such errors often run counter to rules based on
human intuition. However, rules based on human knowledge are challenging to
scale or to even formalize. We thereby seek to infer statistical rules from the
data and quantify the extent to which a model has learned them. We propose a
framework SQRL that integrates logic-based methods with statistical inference
to derive these rules from a model's training data without supervision. We
further show how to adapt models at test time to reduce rule violations and
produce more coherent predictions. SQRL generates up to 300K rules over
datasets from vision, tabular, and language settings. We uncover up to 158K
violations of those rules by state-of-the-art models for classification, object
detection, and data imputation. Test-time adaptation reduces these violations
by up to 68.7% with relative performance improvement up to 32%. SQRL is
available at https://github.com/DebugML/sqrl.
| [
{
"version": "v1",
"created": "Thu, 2 Mar 2023 17:47:02 GMT"
},
{
"version": "v2",
"created": "Tue, 6 Jun 2023 23:22:16 GMT"
}
] | 2023-06-08T00:00:00 | [
[
"Naik",
"Aaditya",
""
],
[
"Wu",
"Yinjun",
""
],
[
"Naik",
"Mayur",
""
],
[
"Wong",
"Eric",
""
]
] | TITLE: Do Machine Learning Models Learn Statistical Rules Inferred from Data?
ABSTRACT: Machine learning models can make critical errors that are easily hidden
within vast amounts of data. Such errors often run counter to rules based on
human intuition. However, rules based on human knowledge are challenging to
scale or to even formalize. We thereby seek to infer statistical rules from the
data and quantify the extent to which a model has learned them. We propose a
framework SQRL that integrates logic-based methods with statistical inference
to derive these rules from a model's training data without supervision. We
further show how to adapt models at test time to reduce rule violations and
produce more coherent predictions. SQRL generates up to 300K rules over
datasets from vision, tabular, and language settings. We uncover up to 158K
violations of those rules by state-of-the-art models for classification, object
detection, and data imputation. Test-time adaptation reduces these violations
by up to 68.7% with relative performance improvement up to 32%. SQRL is
available at https://github.com/DebugML/sqrl.
|
2303.01586 | Qiaozi Gao | Qiaozi Gao, Govind Thattai, Suhaila Shakiah, Xiaofeng Gao, Shreyas
Pansare, Vasu Sharma, Gaurav Sukhatme, Hangjie Shi, Bofei Yang, Desheng
Zheng, Lucy Hu, Karthika Arumugam, Shui Hu, Matthew Wen, Dinakar Guthy,
Cadence Chung, Rohan Khanna, Osman Ipek, Leslie Ball, Kate Bland, Heather
Rocker, Yadunandana Rao, Michael Johnston, Reza Ghanadan, Arindam Mandal,
Dilek Hakkani Tur, Prem Natarajan | Alexa Arena: A User-Centric Interactive Platform for Embodied AI | null | null | null | null | cs.HC cs.AI cs.RO | http://creativecommons.org/licenses/by-nc-sa/4.0/ | We introduce Alexa Arena, a user-centric simulation platform for Embodied AI
(EAI) research. Alexa Arena provides a variety of multi-room layouts and
interactable objects, for the creation of human-robot interaction (HRI)
missions. With user-friendly graphics and control mechanisms, Alexa Arena
supports the development of gamified robotic tasks readily accessible to
general human users, thus opening a new venue for high-efficiency HRI data
collection and EAI system evaluation. Along with the platform, we introduce a
dialog-enabled instruction-following benchmark and provide baseline results for
it. We make Alexa Arena publicly available to facilitate research in building
generalizable and assistive embodied agents.
| [
{
"version": "v1",
"created": "Thu, 2 Mar 2023 21:22:00 GMT"
},
{
"version": "v2",
"created": "Wed, 7 Jun 2023 08:54:46 GMT"
}
] | 2023-06-08T00:00:00 | [
[
"Gao",
"Qiaozi",
""
],
[
"Thattai",
"Govind",
""
],
[
"Shakiah",
"Suhaila",
""
],
[
"Gao",
"Xiaofeng",
""
],
[
"Pansare",
"Shreyas",
""
],
[
"Sharma",
"Vasu",
""
],
[
"Sukhatme",
"Gaurav",
""
],
[
"Shi",
"Hangjie",
""
],
[
"Yang",
"Bofei",
""
],
[
"Zheng",
"Desheng",
""
],
[
"Hu",
"Lucy",
""
],
[
"Arumugam",
"Karthika",
""
],
[
"Hu",
"Shui",
""
],
[
"Wen",
"Matthew",
""
],
[
"Guthy",
"Dinakar",
""
],
[
"Chung",
"Cadence",
""
],
[
"Khanna",
"Rohan",
""
],
[
"Ipek",
"Osman",
""
],
[
"Ball",
"Leslie",
""
],
[
"Bland",
"Kate",
""
],
[
"Rocker",
"Heather",
""
],
[
"Rao",
"Yadunandana",
""
],
[
"Johnston",
"Michael",
""
],
[
"Ghanadan",
"Reza",
""
],
[
"Mandal",
"Arindam",
""
],
[
"Tur",
"Dilek Hakkani",
""
],
[
"Natarajan",
"Prem",
""
]
] | TITLE: Alexa Arena: A User-Centric Interactive Platform for Embodied AI
ABSTRACT: We introduce Alexa Arena, a user-centric simulation platform for Embodied AI
(EAI) research. Alexa Arena provides a variety of multi-room layouts and
interactable objects, for the creation of human-robot interaction (HRI)
missions. With user-friendly graphics and control mechanisms, Alexa Arena
supports the development of gamified robotic tasks readily accessible to
general human users, thus opening a new venue for high-efficiency HRI data
collection and EAI system evaluation. Along with the platform, we introduce a
dialog-enabled instruction-following benchmark and provide baseline results for
it. We make Alexa Arena publicly available to facilitate research in building
generalizable and assistive embodied agents.
|
2303.02535 | Akanksha Saran | Akanksha Saran, Safoora Yousefi, Akshay Krishnamurthy, John Langford,
Jordan T. Ash | Streaming Active Learning with Deep Neural Networks | ICML 2023 | null | null | null | cs.LG | http://creativecommons.org/licenses/by/4.0/ | Active learning is perhaps most naturally posed as an online learning
problem. However, prior active learning approaches with deep neural networks
assume offline access to the entire dataset ahead of time. This paper proposes
VeSSAL, a new algorithm for batch active learning with deep neural networks in
streaming settings, which samples groups of points to query for labels at the
moment they are encountered. Our approach trades off between uncertainty and
diversity of queried samples to match a desired query rate without requiring
any hand-tuned hyperparameters. Altogether, we expand the applicability of deep
neural networks to realistic active learning scenarios, such as applications
relevant to HCI and large, fractured datasets.
| [
{
"version": "v1",
"created": "Sun, 5 Mar 2023 00:57:28 GMT"
},
{
"version": "v2",
"created": "Tue, 6 Jun 2023 23:34:01 GMT"
}
] | 2023-06-08T00:00:00 | [
[
"Saran",
"Akanksha",
""
],
[
"Yousefi",
"Safoora",
""
],
[
"Krishnamurthy",
"Akshay",
""
],
[
"Langford",
"John",
""
],
[
"Ash",
"Jordan T.",
""
]
] | TITLE: Streaming Active Learning with Deep Neural Networks
ABSTRACT: Active learning is perhaps most naturally posed as an online learning
problem. However, prior active learning approaches with deep neural networks
assume offline access to the entire dataset ahead of time. This paper proposes
VeSSAL, a new algorithm for batch active learning with deep neural networks in
streaming settings, which samples groups of points to query for labels at the
moment they are encountered. Our approach trades off between uncertainty and
diversity of queried samples to match a desired query rate without requiring
any hand-tuned hyperparameters. Altogether, we expand the applicability of deep
neural networks to realistic active learning scenarios, such as applications
relevant to HCI and large, fractured datasets.
|
2303.07852 | Bharath Srinivas Prabakaran | Bharath Srinivas Prabakaran and Paul Hamelmann and Erik Ostrowski and
Muhammad Shafique | FPUS23: An Ultrasound Fetus Phantom Dataset with Deep Neural Network
Evaluations for Fetus Orientations, Fetal Planes, and Anatomical Features | Accepted for Publication at IEEE Access | null | 10.1109/ACCESS.2023.3284315 | null | eess.IV cs.CV cs.LG | http://creativecommons.org/licenses/by/4.0/ | Ultrasound imaging is one of the most prominent technologies to evaluate the
growth, progression, and overall health of a fetus during its gestation.
However, the interpretation of the data obtained from such studies is best left
to expert physicians and technicians who are trained and well-versed in
analyzing such images. To improve the clinical workflow and potentially develop
an at-home ultrasound-based fetal monitoring platform, we present a novel fetus
phantom ultrasound dataset, FPUS23, which can be used to identify (1) the
correct diagnostic planes for estimating fetal biometric values, (2) fetus
orientation, (3) their anatomical features, and (4) bounding boxes of the fetus
phantom anatomies at 23 weeks gestation. The entire dataset is composed of
15,728 images, which are used to train four different Deep Neural Network
models, built upon a ResNet34 backbone, for detecting aforementioned fetus
features and use-cases. We have also evaluated the models trained using our
FPUS23 dataset, to show that the information learned by these models can be
used to substantially increase the accuracy on real-world ultrasound fetus
datasets. We make the FPUS23 dataset and the pre-trained models publicly
accessible at https://github.com/bharathprabakaran/FPUS23, which will further
facilitate future research on fetal ultrasound imaging and analysis.
| [
{
"version": "v1",
"created": "Tue, 14 Mar 2023 12:46:48 GMT"
},
{
"version": "v2",
"created": "Wed, 7 Jun 2023 14:56:10 GMT"
}
] | 2023-06-08T00:00:00 | [
[
"Prabakaran",
"Bharath Srinivas",
""
],
[
"Hamelmann",
"Paul",
""
],
[
"Ostrowski",
"Erik",
""
],
[
"Shafique",
"Muhammad",
""
]
] | TITLE: FPUS23: An Ultrasound Fetus Phantom Dataset with Deep Neural Network
Evaluations for Fetus Orientations, Fetal Planes, and Anatomical Features
ABSTRACT: Ultrasound imaging is one of the most prominent technologies to evaluate the
growth, progression, and overall health of a fetus during its gestation.
However, the interpretation of the data obtained from such studies is best left
to expert physicians and technicians who are trained and well-versed in
analyzing such images. To improve the clinical workflow and potentially develop
an at-home ultrasound-based fetal monitoring platform, we present a novel fetus
phantom ultrasound dataset, FPUS23, which can be used to identify (1) the
correct diagnostic planes for estimating fetal biometric values, (2) fetus
orientation, (3) their anatomical features, and (4) bounding boxes of the fetus
phantom anatomies at 23 weeks gestation. The entire dataset is composed of
15,728 images, which are used to train four different Deep Neural Network
models, built upon a ResNet34 backbone, for detecting aforementioned fetus
features and use-cases. We have also evaluated the models trained using our
FPUS23 dataset, to show that the information learned by these models can be
used to substantially increase the accuracy on real-world ultrasound fetus
datasets. We make the FPUS23 dataset and the pre-trained models publicly
accessible at https://github.com/bharathprabakaran/FPUS23, which will further
facilitate future research on fetal ultrasound imaging and analysis.
|
2303.11926 | Yingfei Liu | Shihao Wang, Yingfei Liu, Tiancai Wang, Ying Li, Xiangyu Zhang | Exploring Object-Centric Temporal Modeling for Efficient Multi-View 3D
Object Detection | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this paper, we propose a long-sequence modeling framework, named
StreamPETR, for multi-view 3D object detection. Built upon the sparse query
design in the PETR series, we systematically develop an object-centric temporal
mechanism. The model is performed in an online manner and the long-term
historical information is propagated through object queries frame by frame.
Besides, we introduce a motion-aware layer normalization to model the movement
of the objects. StreamPETR achieves significant performance improvements only
with negligible computation cost, compared to the single-frame baseline. On the
standard nuScenes benchmark, it is the first online multi-view method that
achieves comparable performance (67.6% NDS & 65.3% AMOTA) with lidar-based
methods. The lightweight version realizes 45.0% mAP and 31.7 FPS, outperforming
the state-of-the-art method (SOLOFusion) by 2.3% mAP and 1.8x faster FPS. Code
has been available at https://github.com/exiawsh/StreamPETR.git.
| [
{
"version": "v1",
"created": "Tue, 21 Mar 2023 15:19:20 GMT"
},
{
"version": "v2",
"created": "Wed, 7 Jun 2023 08:08:16 GMT"
}
] | 2023-06-08T00:00:00 | [
[
"Wang",
"Shihao",
""
],
[
"Liu",
"Yingfei",
""
],
[
"Wang",
"Tiancai",
""
],
[
"Li",
"Ying",
""
],
[
"Zhang",
"Xiangyu",
""
]
] | TITLE: Exploring Object-Centric Temporal Modeling for Efficient Multi-View 3D
Object Detection
ABSTRACT: In this paper, we propose a long-sequence modeling framework, named
StreamPETR, for multi-view 3D object detection. Built upon the sparse query
design in the PETR series, we systematically develop an object-centric temporal
mechanism. The model is performed in an online manner and the long-term
historical information is propagated through object queries frame by frame.
Besides, we introduce a motion-aware layer normalization to model the movement
of the objects. StreamPETR achieves significant performance improvements only
with negligible computation cost, compared to the single-frame baseline. On the
standard nuScenes benchmark, it is the first online multi-view method that
achieves comparable performance (67.6% NDS & 65.3% AMOTA) with lidar-based
methods. The lightweight version realizes 45.0% mAP and 31.7 FPS, outperforming
the state-of-the-art method (SOLOFusion) by 2.3% mAP and 1.8x faster FPS. Code
has been available at https://github.com/exiawsh/StreamPETR.git.
|
2303.13933 | Ye Mao | Ye Mao, Lan Jiang, Xi Chen, and Chao Li | DisC-Diff: Disentangled Conditional Diffusion Model for Multi-Contrast
MRI Super-Resolution | Early Accepted by MICCAI 2023 | null | null | null | eess.IV cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Multi-contrast magnetic resonance imaging (MRI) is the most common management
tool used to characterize neurological disorders based on brain tissue
contrasts. However, acquiring high-resolution MRI scans is time-consuming and
infeasible under specific conditions. Hence, multi-contrast super-resolution
methods have been developed to improve the quality of low-resolution contrasts
by leveraging complementary information from multi-contrast MRI. Current deep
learning-based super-resolution methods have limitations in estimating
restoration uncertainty and avoiding mode collapse. Although the diffusion
model has emerged as a promising approach for image enhancement, capturing
complex interactions between multiple conditions introduced by multi-contrast
MRI super-resolution remains a challenge for clinical applications. In this
paper, we propose a disentangled conditional diffusion model, DisC-Diff, for
multi-contrast brain MRI super-resolution. It utilizes the sampling-based
generation and simple objective function of diffusion models to estimate
uncertainty in restorations effectively and ensure a stable optimization
process. Moreover, DisC-Diff leverages a disentangled multi-stream network to
fully exploit complementary information from multi-contrast MRI, improving
model interpretation under multiple conditions of multi-contrast inputs. We
validated the effectiveness of DisC-Diff on two datasets: the IXI dataset,
which contains 578 normal brains, and a clinical dataset with 316 pathological
brains. Our experimental results demonstrate that DisC-Diff outperforms other
state-of-the-art methods both quantitatively and visually.
| [
{
"version": "v1",
"created": "Fri, 24 Mar 2023 11:42:45 GMT"
},
{
"version": "v2",
"created": "Tue, 6 Jun 2023 20:22:06 GMT"
}
] | 2023-06-08T00:00:00 | [
[
"Mao",
"Ye",
""
],
[
"Jiang",
"Lan",
""
],
[
"Chen",
"Xi",
""
],
[
"Li",
"Chao",
""
]
] | TITLE: DisC-Diff: Disentangled Conditional Diffusion Model for Multi-Contrast
MRI Super-Resolution
ABSTRACT: Multi-contrast magnetic resonance imaging (MRI) is the most common management
tool used to characterize neurological disorders based on brain tissue
contrasts. However, acquiring high-resolution MRI scans is time-consuming and
infeasible under specific conditions. Hence, multi-contrast super-resolution
methods have been developed to improve the quality of low-resolution contrasts
by leveraging complementary information from multi-contrast MRI. Current deep
learning-based super-resolution methods have limitations in estimating
restoration uncertainty and avoiding mode collapse. Although the diffusion
model has emerged as a promising approach for image enhancement, capturing
complex interactions between multiple conditions introduced by multi-contrast
MRI super-resolution remains a challenge for clinical applications. In this
paper, we propose a disentangled conditional diffusion model, DisC-Diff, for
multi-contrast brain MRI super-resolution. It utilizes the sampling-based
generation and simple objective function of diffusion models to estimate
uncertainty in restorations effectively and ensure a stable optimization
process. Moreover, DisC-Diff leverages a disentangled multi-stream network to
fully exploit complementary information from multi-contrast MRI, improving
model interpretation under multiple conditions of multi-contrast inputs. We
validated the effectiveness of DisC-Diff on two datasets: the IXI dataset,
which contains 578 normal brains, and a clinical dataset with 316 pathological
brains. Our experimental results demonstrate that DisC-Diff outperforms other
state-of-the-art methods both quantitatively and visually.
|
2303.14068 | Md Asif Bin Syed | Md Asif Bin Syed and Imtiaz Ahmed | A CNN-LSTM Architecture for Marine Vessel Track Association Using
Automatic Identification System (AIS) Data | No results has been changed, Changed Figures, increased the
resolution of the figures, changed slightly in the description in section
3.3.3 | null | null | null | cs.LG | http://creativecommons.org/licenses/by-nc-nd/4.0/ | In marine surveillance, distinguishing between normal and anomalous vessel
movement patterns is critical for identifying potential threats in a timely
manner. Once detected, it is important to monitor and track these vessels until
a necessary intervention occurs. To achieve this, track association algorithms
are used, which take sequential observations comprising geological and motion
parameters of the vessels and associate them with respective vessels. The
spatial and temporal variations inherent in these sequential observations make
the association task challenging for traditional multi-object tracking
algorithms. Additionally, the presence of overlapping tracks and missing data
can further complicate the trajectory tracking process. To address these
challenges, in this study, we approach this tracking task as a multivariate
time series problem and introduce a 1D CNN-LSTM architecture-based framework
for track association. This special neural network architecture can capture the
spatial patterns as well as the long-term temporal relations that exist among
the sequential observations. During the training process, it learns and builds
the trajectory for each of these underlying vessels. Once trained, the proposed
framework takes the marine vessel's location and motion data collected through
the Automatic Identification System (AIS) as input and returns the most likely
vessel track as output in real-time. To evaluate the performance of our
approach, we utilize an AIS dataset containing observations from 327 vessels
traveling in a specific geographic region. We measure the performance of our
proposed framework using standard performance metrics such as accuracy,
precision, recall, and F1 score. When compared with other competitive neural
network architectures our approach demonstrates a superior tracking
performance.
| [
{
"version": "v1",
"created": "Fri, 24 Mar 2023 15:26:49 GMT"
},
{
"version": "v2",
"created": "Wed, 7 Jun 2023 00:23:57 GMT"
}
] | 2023-06-08T00:00:00 | [
[
"Syed",
"Md Asif Bin",
""
],
[
"Ahmed",
"Imtiaz",
""
]
] | TITLE: A CNN-LSTM Architecture for Marine Vessel Track Association Using
Automatic Identification System (AIS) Data
ABSTRACT: In marine surveillance, distinguishing between normal and anomalous vessel
movement patterns is critical for identifying potential threats in a timely
manner. Once detected, it is important to monitor and track these vessels until
a necessary intervention occurs. To achieve this, track association algorithms
are used, which take sequential observations comprising geological and motion
parameters of the vessels and associate them with respective vessels. The
spatial and temporal variations inherent in these sequential observations make
the association task challenging for traditional multi-object tracking
algorithms. Additionally, the presence of overlapping tracks and missing data
can further complicate the trajectory tracking process. To address these
challenges, in this study, we approach this tracking task as a multivariate
time series problem and introduce a 1D CNN-LSTM architecture-based framework
for track association. This special neural network architecture can capture the
spatial patterns as well as the long-term temporal relations that exist among
the sequential observations. During the training process, it learns and builds
the trajectory for each of these underlying vessels. Once trained, the proposed
framework takes the marine vessel's location and motion data collected through
the Automatic Identification System (AIS) as input and returns the most likely
vessel track as output in real-time. To evaluate the performance of our
approach, we utilize an AIS dataset containing observations from 327 vessels
traveling in a specific geographic region. We measure the performance of our
proposed framework using standard performance metrics such as accuracy,
precision, recall, and F1 score. When compared with other competitive neural
network architectures our approach demonstrates a superior tracking
performance.
|
2304.06009 | Alexander Naumann | Alexander Naumann, Felix Hertlein, Laura D\"orr, Steffen Thoma, Kai
Furmans | Literature Review: Computer Vision Applications in Transportation
Logistics and Warehousing | null | null | null | null | cs.CV cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Computer vision applications in transportation logistics and warehousing have
a huge potential for process automation. We present a structured literature
review on research in the field to help leverage this potential. The literature
is categorized w.r.t. the application, i.e. the task it tackles and w.r.t. the
computer vision techniques that are used. Regarding applications, we subdivide
the literature in two areas: Monitoring, i.e. observing and retrieving relevant
information from the environment, and manipulation, where approaches are used
to analyze and interact with the environment. Additionally, we point out
directions for future research and link to recent developments in computer
vision that are suitable for application in logistics. Finally, we present an
overview of existing datasets and industrial solutions. The results of our
analysis are also available online at https://a-nau.github.io/cv-in-logistics.
| [
{
"version": "v1",
"created": "Wed, 12 Apr 2023 17:33:41 GMT"
},
{
"version": "v2",
"created": "Wed, 7 Jun 2023 10:05:59 GMT"
}
] | 2023-06-08T00:00:00 | [
[
"Naumann",
"Alexander",
""
],
[
"Hertlein",
"Felix",
""
],
[
"Dörr",
"Laura",
""
],
[
"Thoma",
"Steffen",
""
],
[
"Furmans",
"Kai",
""
]
] | TITLE: Literature Review: Computer Vision Applications in Transportation
Logistics and Warehousing
ABSTRACT: Computer vision applications in transportation logistics and warehousing have
a huge potential for process automation. We present a structured literature
review on research in the field to help leverage this potential. The literature
is categorized w.r.t. the application, i.e. the task it tackles and w.r.t. the
computer vision techniques that are used. Regarding applications, we subdivide
the literature in two areas: Monitoring, i.e. observing and retrieving relevant
information from the environment, and manipulation, where approaches are used
to analyze and interact with the environment. Additionally, we point out
directions for future research and link to recent developments in computer
vision that are suitable for application in logistics. Finally, we present an
overview of existing datasets and industrial solutions. The results of our
analysis are also available online at https://a-nau.github.io/cv-in-logistics.
|
2304.06051 | Yifeng Shi | Yifeng Shi and Feng Lv and Xinliang Wang and Chunlong Xia and Shaojie
Li and Shujie Yang and Teng Xi and Gang Zhang | Open-TransMind: A New Baseline and Benchmark for 1st Foundation Model
Challenge of Intelligent Transportation | null | null | null | null | cs.CV cs.AI cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | With the continuous improvement of computing power and deep learning
algorithms in recent years, the foundation model has grown in popularity.
Because of its powerful capabilities and excellent performance, this technology
is being adopted and applied by an increasing number of industries. In the
intelligent transportation industry, artificial intelligence faces the
following typical challenges: few shots, poor generalization, and a lack of
multi-modal techniques. Foundation model technology can significantly alleviate
the aforementioned issues. To address these, we designed the 1st Foundation
Model Challenge, with the goal of increasing the popularity of foundation model
technology in traffic scenarios and promoting the rapid development of the
intelligent transportation industry. The challenge is divided into two tracks:
all-in-one and cross-modal image retrieval. Furthermore, we provide a new
baseline and benchmark for the two tracks, called Open-TransMind. According to
our knowledge, Open-TransMind is the first open-source transportation
foundation model with multi-task and multi-modal capabilities. Simultaneously,
Open-TransMind can achieve state-of-the-art performance on detection,
classification, and segmentation datasets of traffic scenarios. Our source code
is available at https://github.com/Traffic-X/Open-TransMind.
| [
{
"version": "v1",
"created": "Wed, 12 Apr 2023 04:27:16 GMT"
},
{
"version": "v2",
"created": "Wed, 7 Jun 2023 12:29:25 GMT"
}
] | 2023-06-08T00:00:00 | [
[
"Shi",
"Yifeng",
""
],
[
"Lv",
"Feng",
""
],
[
"Wang",
"Xinliang",
""
],
[
"Xia",
"Chunlong",
""
],
[
"Li",
"Shaojie",
""
],
[
"Yang",
"Shujie",
""
],
[
"Xi",
"Teng",
""
],
[
"Zhang",
"Gang",
""
]
] | TITLE: Open-TransMind: A New Baseline and Benchmark for 1st Foundation Model
Challenge of Intelligent Transportation
ABSTRACT: With the continuous improvement of computing power and deep learning
algorithms in recent years, the foundation model has grown in popularity.
Because of its powerful capabilities and excellent performance, this technology
is being adopted and applied by an increasing number of industries. In the
intelligent transportation industry, artificial intelligence faces the
following typical challenges: few shots, poor generalization, and a lack of
multi-modal techniques. Foundation model technology can significantly alleviate
the aforementioned issues. To address these, we designed the 1st Foundation
Model Challenge, with the goal of increasing the popularity of foundation model
technology in traffic scenarios and promoting the rapid development of the
intelligent transportation industry. The challenge is divided into two tracks:
all-in-one and cross-modal image retrieval. Furthermore, we provide a new
baseline and benchmark for the two tracks, called Open-TransMind. According to
our knowledge, Open-TransMind is the first open-source transportation
foundation model with multi-task and multi-modal capabilities. Simultaneously,
Open-TransMind can achieve state-of-the-art performance on detection,
classification, and segmentation datasets of traffic scenarios. Our source code
is available at https://github.com/Traffic-X/Open-TransMind.
|
2304.13840 | Enrique Dehaerne | Enrique Dehaerne and Bappaditya Dey and Sandip Halder and Stefan De
Gendt | A Deep Learning Framework for Verilog Autocompletion Towards Design and
Verification Automation | Updated text to correct language errors and added a link to
supplementary code and data
(https://github.com/99EnriqueD/verilog_autocompletion). 6 pages, 3 figures, 4
tables. To be presented as a WIP poster at DAC 2023 | null | null | null | cs.LG cs.SE | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Innovative Electronic Design Automation (EDA) solutions are important to meet
the design requirements for increasingly complex electronic devices. Verilog, a
hardware description language, is widely used for the design and verification
of digital circuits and is synthesized using specific EDA tools. However,
writing code is a repetitive and time-intensive task. This paper proposes,
primarily, a novel deep learning framework for training a Verilog
autocompletion model and, secondarily, a Verilog dataset of files and snippets
obtained from open-source repositories. The framework involves integrating
models pretrained on general programming language data and finetuning them on a
dataset curated to be similar to a target downstream task. This is validated by
comparing different pretrained models trained on different subsets of the
proposed Verilog dataset using multiple evaluation metrics. These experiments
demonstrate that the proposed framework achieves better BLEU, ROUGE-L, and chrF
scores by 9.5%, 6.7%, and 6.9%, respectively, compared to a model trained from
scratch. Code and data are made available at:
https://github.com/99EnriqueD/verilog_autocompletion .
| [
{
"version": "v1",
"created": "Wed, 26 Apr 2023 21:56:03 GMT"
},
{
"version": "v2",
"created": "Wed, 7 Jun 2023 09:33:04 GMT"
}
] | 2023-06-08T00:00:00 | [
[
"Dehaerne",
"Enrique",
""
],
[
"Dey",
"Bappaditya",
""
],
[
"Halder",
"Sandip",
""
],
[
"De Gendt",
"Stefan",
""
]
] | TITLE: A Deep Learning Framework for Verilog Autocompletion Towards Design and
Verification Automation
ABSTRACT: Innovative Electronic Design Automation (EDA) solutions are important to meet
the design requirements for increasingly complex electronic devices. Verilog, a
hardware description language, is widely used for the design and verification
of digital circuits and is synthesized using specific EDA tools. However,
writing code is a repetitive and time-intensive task. This paper proposes,
primarily, a novel deep learning framework for training a Verilog
autocompletion model and, secondarily, a Verilog dataset of files and snippets
obtained from open-source repositories. The framework involves integrating
models pretrained on general programming language data and finetuning them on a
dataset curated to be similar to a target downstream task. This is validated by
comparing different pretrained models trained on different subsets of the
proposed Verilog dataset using multiple evaluation metrics. These experiments
demonstrate that the proposed framework achieves better BLEU, ROUGE-L, and chrF
scores by 9.5%, 6.7%, and 6.9%, respectively, compared to a model trained from
scratch. Code and data are made available at:
https://github.com/99EnriqueD/verilog_autocompletion .
|
2304.14400 | Ronghuan Wu | Ronghuan Wu, Wanchao Su, Kede Ma, Jing Liao | IconShop: Text-Guided Vector Icon Synthesis with Autoregressive
Transformers | Project Page: https://icon-shop.github.io/ | null | null | null | cs.CV cs.GR | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Scalable Vector Graphics (SVG) is a popular vector image format that offers
good support for interactivity and animation. Despite its appealing
characteristics, creating custom SVG content can be challenging for users due
to the steep learning curve required to understand SVG grammars or get familiar
with professional editing software. Recent advancements in text-to-image
generation have inspired researchers to explore vector graphics synthesis using
either image-based methods (i.e., text -> raster image -> vector graphics)
combining text-to-image generation models with image vectorization, or
language-based methods (i.e., text -> vector graphics script) through
pretrained large language models. However, these methods still suffer from
limitations in terms of generation quality, diversity, and flexibility. In this
paper, we introduce IconShop, a text-guided vector icon synthesis method using
autoregressive transformers. The key to success of our approach is to
sequentialize and tokenize SVG paths (and textual descriptions as guidance)
into a uniquely decodable token sequence. With that, we are able to fully
exploit the sequence learning power of autoregressive transformers, while
enabling both unconditional and text-conditioned icon synthesis. Through
standard training to predict the next token on a large-scale vector icon
dataset accompanied by textural descriptions, the proposed IconShop
consistently exhibits better icon synthesis capability than existing
image-based and language-based methods both quantitatively and qualitatively.
Meanwhile, we observe a dramatic improvement in generation diversity, which is
validated by the objective Uniqueness and Novelty measures. More importantly,
we demonstrate the flexibility of IconShop with multiple novel icon synthesis
tasks, including icon editing, icon interpolation, icon semantic combination,
and icon design auto-suggestion.
| [
{
"version": "v1",
"created": "Thu, 27 Apr 2023 17:58:02 GMT"
},
{
"version": "v2",
"created": "Wed, 24 May 2023 13:45:42 GMT"
},
{
"version": "v3",
"created": "Mon, 29 May 2023 14:26:12 GMT"
},
{
"version": "v4",
"created": "Wed, 7 Jun 2023 03:45:20 GMT"
}
] | 2023-06-08T00:00:00 | [
[
"Wu",
"Ronghuan",
""
],
[
"Su",
"Wanchao",
""
],
[
"Ma",
"Kede",
""
],
[
"Liao",
"Jing",
""
]
] | TITLE: IconShop: Text-Guided Vector Icon Synthesis with Autoregressive
Transformers
ABSTRACT: Scalable Vector Graphics (SVG) is a popular vector image format that offers
good support for interactivity and animation. Despite its appealing
characteristics, creating custom SVG content can be challenging for users due
to the steep learning curve required to understand SVG grammars or get familiar
with professional editing software. Recent advancements in text-to-image
generation have inspired researchers to explore vector graphics synthesis using
either image-based methods (i.e., text -> raster image -> vector graphics)
combining text-to-image generation models with image vectorization, or
language-based methods (i.e., text -> vector graphics script) through
pretrained large language models. However, these methods still suffer from
limitations in terms of generation quality, diversity, and flexibility. In this
paper, we introduce IconShop, a text-guided vector icon synthesis method using
autoregressive transformers. The key to success of our approach is to
sequentialize and tokenize SVG paths (and textual descriptions as guidance)
into a uniquely decodable token sequence. With that, we are able to fully
exploit the sequence learning power of autoregressive transformers, while
enabling both unconditional and text-conditioned icon synthesis. Through
standard training to predict the next token on a large-scale vector icon
dataset accompanied by textural descriptions, the proposed IconShop
consistently exhibits better icon synthesis capability than existing
image-based and language-based methods both quantitatively and qualitatively.
Meanwhile, we observe a dramatic improvement in generation diversity, which is
validated by the objective Uniqueness and Novelty measures. More importantly,
we demonstrate the flexibility of IconShop with multiple novel icon synthesis
tasks, including icon editing, icon interpolation, icon semantic combination,
and icon design auto-suggestion.
|
2305.04294 | Zhen Huang | Zhen Huang, Han Li, Shitong Shao, Heqin Zhu, Huijie Hu, Zhiwei Cheng,
Jianji Wang, and S.Kevin Zhou | PELE scores: Pelvic X-ray Landmark Detection by Pelvis Extraction and
Enhancement | will revise it and resubmit it again later | null | null | null | eess.IV cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The pelvis, the lower part of the trunk, supports and balances the trunk.
Landmark detection from a pelvic X-ray (PXR) facilitates downstream analysis
and computer-assisted diagnosis and treatment of pelvic diseases. Although PXRs
have the advantages of low radiation and reduced cost compared to computed
tomography (CT) images, their 2D pelvis-tissue superposition of 3D structures
confuses clinical decision-making. In this paper, we propose a PELvis
Extraction (PELE) module that utilizes 3D prior anatomical knowledge in CT to
guide and well isolate the pelvis from PXRs, thereby eliminating the influence
of soft tissue. We conduct an extensive evaluation based on two public datasets
and one private dataset, totaling 850 PXRs. The experimental results show that
the proposed PELE module significantly improves the accuracy of PXRs landmark
detection and achieves state-of-the-art performances in several benchmark
metrics, thus better serving downstream tasks.
| [
{
"version": "v1",
"created": "Sun, 7 May 2023 14:45:04 GMT"
},
{
"version": "v2",
"created": "Wed, 7 Jun 2023 04:59:12 GMT"
}
] | 2023-06-08T00:00:00 | [
[
"Huang",
"Zhen",
""
],
[
"Li",
"Han",
""
],
[
"Shao",
"Shitong",
""
],
[
"Zhu",
"Heqin",
""
],
[
"Hu",
"Huijie",
""
],
[
"Cheng",
"Zhiwei",
""
],
[
"Wang",
"Jianji",
""
],
[
"Zhou",
"S. Kevin",
""
]
] | TITLE: PELE scores: Pelvic X-ray Landmark Detection by Pelvis Extraction and
Enhancement
ABSTRACT: The pelvis, the lower part of the trunk, supports and balances the trunk.
Landmark detection from a pelvic X-ray (PXR) facilitates downstream analysis
and computer-assisted diagnosis and treatment of pelvic diseases. Although PXRs
have the advantages of low radiation and reduced cost compared to computed
tomography (CT) images, their 2D pelvis-tissue superposition of 3D structures
confuses clinical decision-making. In this paper, we propose a PELvis
Extraction (PELE) module that utilizes 3D prior anatomical knowledge in CT to
guide and well isolate the pelvis from PXRs, thereby eliminating the influence
of soft tissue. We conduct an extensive evaluation based on two public datasets
and one private dataset, totaling 850 PXRs. The experimental results show that
the proposed PELE module significantly improves the accuracy of PXRs landmark
detection and achieves state-of-the-art performances in several benchmark
metrics, thus better serving downstream tasks.
|
2305.09397 | Anuj Rai | Anuj Rai, Somnath Dey | EXPRESSNET: An Explainable Residual Slim Network for Fingerprint
Presentation Attack Detection | arXiv admin note: text overlap with arXiv:2303.01465 | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | Presentation attack is a challenging issue that persists in the security of
automatic fingerprint recognition systems. This paper proposes a novel
explainable residual slim network that detects the presentation attack by
representing the visual features in the input fingerprint sample. The
encoder-decoder of this network along with the channel attention block converts
the input sample into its heatmap representation while the modified residual
convolutional neural network classifier discriminates between live and spoof
fingerprints. The entire architecture of the heatmap generator block and
modified ResNet classifier works together in an end-to-end manner. The
performance of the proposed model is validated on benchmark liveness detection
competition databases i.e. Livdet 2011, 2013, 2015, 2017, and 2019 and the
classification accuracy of 96.86\%, 99.84\%, 96.45\%, 96.07\%, 96.27\% are
achieved on them, respectively. The performance of the proposed model is
compared with the state-of-the-art techniques, and the proposed method
outperforms state-of-the-art methods in benchmark protocols of presentation
attack detection in terms of classification accuracy.
| [
{
"version": "v1",
"created": "Tue, 16 May 2023 12:29:50 GMT"
},
{
"version": "v2",
"created": "Tue, 6 Jun 2023 18:24:21 GMT"
}
] | 2023-06-08T00:00:00 | [
[
"Rai",
"Anuj",
""
],
[
"Dey",
"Somnath",
""
]
] | TITLE: EXPRESSNET: An Explainable Residual Slim Network for Fingerprint
Presentation Attack Detection
ABSTRACT: Presentation attack is a challenging issue that persists in the security of
automatic fingerprint recognition systems. This paper proposes a novel
explainable residual slim network that detects the presentation attack by
representing the visual features in the input fingerprint sample. The
encoder-decoder of this network along with the channel attention block converts
the input sample into its heatmap representation while the modified residual
convolutional neural network classifier discriminates between live and spoof
fingerprints. The entire architecture of the heatmap generator block and
modified ResNet classifier works together in an end-to-end manner. The
performance of the proposed model is validated on benchmark liveness detection
competition databases i.e. Livdet 2011, 2013, 2015, 2017, and 2019 and the
classification accuracy of 96.86\%, 99.84\%, 96.45\%, 96.07\%, 96.27\% are
achieved on them, respectively. The performance of the proposed model is
compared with the state-of-the-art techniques, and the proposed method
outperforms state-of-the-art methods in benchmark protocols of presentation
attack detection in terms of classification accuracy.
|
2305.09674 | Gr\'egoire Barru\'e | Gr\'egoire Barru\'e and Tony Quertier | Quantum Machine Learning for Malware Classification | null | null | null | null | cs.CR cs.ET cs.LG quant-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In a context of malicious software detection, machine learning (ML) is widely
used to generalize to new malware. However, it has been demonstrated that ML
models can be fooled or may have generalization problems on malware that has
never been seen. We investigate the possible benefits of quantum algorithms for
classification tasks. We implement two models of Quantum Machine Learning
algorithms, and we compare them to classical models for the classification of a
dataset composed of malicious and benign executable files. We try to optimize
our algorithms based on methods found in the literature, and analyze our
results in an exploratory way, to identify the most interesting directions to
explore for the future.
| [
{
"version": "v1",
"created": "Tue, 9 May 2023 09:21:48 GMT"
},
{
"version": "v2",
"created": "Mon, 22 May 2023 07:38:57 GMT"
},
{
"version": "v3",
"created": "Wed, 7 Jun 2023 07:56:11 GMT"
}
] | 2023-06-08T00:00:00 | [
[
"Barrué",
"Grégoire",
""
],
[
"Quertier",
"Tony",
""
]
] | TITLE: Quantum Machine Learning for Malware Classification
ABSTRACT: In a context of malicious software detection, machine learning (ML) is widely
used to generalize to new malware. However, it has been demonstrated that ML
models can be fooled or may have generalization problems on malware that has
never been seen. We investigate the possible benefits of quantum algorithms for
classification tasks. We implement two models of Quantum Machine Learning
algorithms, and we compare them to classical models for the classification of a
dataset composed of malicious and benign executable files. We try to optimize
our algorithms based on methods found in the literature, and analyze our
results in an exploratory way, to identify the most interesting directions to
explore for the future.
|
2305.14070 | Leonardo Zilio | Rodrigo Wilkens, Leonardo Zilio and Aline Villavicencio | Assessing Linguistic Generalisation in Language Models: A Dataset for
Brazilian Portuguese | This is the original manuscript that was submitted to LREV. The final
version was published recently and can be found at: https://rdcu.be/ddEa6.
Language Resources and Evaluation, https://doi.org/10.1007/s10579-023-09664-1 | null | null | null | cs.CL | http://creativecommons.org/licenses/by-sa/4.0/ | Much recent effort has been devoted to creating large-scale language models.
Nowadays, the most prominent approaches are based on deep neural networks, such
as BERT. However, they lack transparency and interpretability, and are often
seen as black boxes. This affects not only their applicability in downstream
tasks but also the comparability of different architectures or even of the same
model trained using different corpora or hyperparameters. In this paper, we
propose a set of intrinsic evaluation tasks that inspect the linguistic
information encoded in models developed for Brazilian Portuguese. These tasks
are designed to evaluate how different language models generalise information
related to grammatical structures and multiword expressions (MWEs), thus
allowing for an assessment of whether the model has learned different
linguistic phenomena. The dataset that was developed for these tasks is
composed of a series of sentences with a single masked word and a cue phrase
that helps in narrowing down the context. This dataset is divided into MWEs and
grammatical structures, and the latter is subdivided into 6 tasks: impersonal
verbs, subject agreement, verb agreement, nominal agreement, passive and
connectors. The subset for MWEs was used to test BERTimbau Large, BERTimbau
Base and mBERT. For the grammatical structures, we used only BERTimbau Large,
because it yielded the best results in the MWE task.
| [
{
"version": "v1",
"created": "Tue, 23 May 2023 13:49:14 GMT"
},
{
"version": "v2",
"created": "Wed, 7 Jun 2023 08:53:08 GMT"
}
] | 2023-06-08T00:00:00 | [
[
"Wilkens",
"Rodrigo",
""
],
[
"Zilio",
"Leonardo",
""
],
[
"Villavicencio",
"Aline",
""
]
] | TITLE: Assessing Linguistic Generalisation in Language Models: A Dataset for
Brazilian Portuguese
ABSTRACT: Much recent effort has been devoted to creating large-scale language models.
Nowadays, the most prominent approaches are based on deep neural networks, such
as BERT. However, they lack transparency and interpretability, and are often
seen as black boxes. This affects not only their applicability in downstream
tasks but also the comparability of different architectures or even of the same
model trained using different corpora or hyperparameters. In this paper, we
propose a set of intrinsic evaluation tasks that inspect the linguistic
information encoded in models developed for Brazilian Portuguese. These tasks
are designed to evaluate how different language models generalise information
related to grammatical structures and multiword expressions (MWEs), thus
allowing for an assessment of whether the model has learned different
linguistic phenomena. The dataset that was developed for these tasks is
composed of a series of sentences with a single masked word and a cue phrase
that helps in narrowing down the context. This dataset is divided into MWEs and
grammatical structures, and the latter is subdivided into 6 tasks: impersonal
verbs, subject agreement, verb agreement, nominal agreement, passive and
connectors. The subset for MWEs was used to test BERTimbau Large, BERTimbau
Base and mBERT. For the grammatical structures, we used only BERTimbau Large,
because it yielded the best results in the MWE task.
|
2305.16253 | Yan Liu | Yan Liu, Yan Gao, Zhe Su, Xiaokang Chen, Elliott Ash, Jian-Guang Lou | Uncovering and Categorizing Social Biases in Text-to-SQL | null | null | null | null | cs.CL | http://creativecommons.org/licenses/by/4.0/ | Content Warning: This work contains examples that potentially implicate
stereotypes, associations, and other harms that could be offensive to
individuals in certain social groups.} Large pre-trained language models are
acknowledged to carry social biases towards different demographics, which can
further amplify existing stereotypes in our society and cause even more harm.
Text-to-SQL is an important task, models of which are mainly adopted by
administrative industries, where unfair decisions may lead to catastrophic
consequences. However, existing Text-to-SQL models are trained on clean,
neutral datasets, such as Spider and WikiSQL. This, to some extent, cover up
social bias in models under ideal conditions, which nevertheless may emerge in
real application scenarios. In this work, we aim to uncover and categorize
social biases in Text-to-SQL models. We summarize the categories of social
biases that may occur in structured data for Text-to-SQL models. We build test
benchmarks and reveal that models with similar task accuracy can contain social
biases at very different rates. We show how to take advantage of our
methodology to uncover and assess social biases in the downstream Text-to-SQL
task. We will release our code and data.
| [
{
"version": "v1",
"created": "Thu, 25 May 2023 17:08:56 GMT"
},
{
"version": "v2",
"created": "Wed, 7 Jun 2023 13:30:39 GMT"
}
] | 2023-06-08T00:00:00 | [
[
"Liu",
"Yan",
""
],
[
"Gao",
"Yan",
""
],
[
"Su",
"Zhe",
""
],
[
"Chen",
"Xiaokang",
""
],
[
"Ash",
"Elliott",
""
],
[
"Lou",
"Jian-Guang",
""
]
] | TITLE: Uncovering and Categorizing Social Biases in Text-to-SQL
ABSTRACT: Content Warning: This work contains examples that potentially implicate
stereotypes, associations, and other harms that could be offensive to
individuals in certain social groups.} Large pre-trained language models are
acknowledged to carry social biases towards different demographics, which can
further amplify existing stereotypes in our society and cause even more harm.
Text-to-SQL is an important task, models of which are mainly adopted by
administrative industries, where unfair decisions may lead to catastrophic
consequences. However, existing Text-to-SQL models are trained on clean,
neutral datasets, such as Spider and WikiSQL. This, to some extent, cover up
social bias in models under ideal conditions, which nevertheless may emerge in
real application scenarios. In this work, we aim to uncover and categorize
social biases in Text-to-SQL models. We summarize the categories of social
biases that may occur in structured data for Text-to-SQL models. We build test
benchmarks and reveal that models with similar task accuracy can contain social
biases at very different rates. We show how to take advantage of our
methodology to uncover and assess social biases in the downstream Text-to-SQL
task. We will release our code and data.
|
2305.16636 | Vijay Viswanathan | Vijay Viswanathan, Luyu Gao, Tongshuang Wu, Pengfei Liu and Graham
Neubig | DataFinder: Scientific Dataset Recommendation from Natural Language
Descriptions | To appear at ACL 2023. Code published at
https://github.com/viswavi/datafinder | null | null | null | cs.IR cs.CL cs.DL | http://creativecommons.org/licenses/by/4.0/ | Modern machine learning relies on datasets to develop and validate research
ideas. Given the growth of publicly available data, finding the right dataset
to use is increasingly difficult. Any research question imposes explicit and
implicit constraints on how well a given dataset will enable researchers to
answer this question, such as dataset size, modality, and domain. We
operationalize the task of recommending datasets given a short natural language
description of a research idea, to help people find relevant datasets for their
needs. Dataset recommendation poses unique challenges as an information
retrieval problem; datasets are hard to directly index for search and there are
no corpora readily available for this task. To facilitate this task, we build
the DataFinder Dataset which consists of a larger automatically-constructed
training set (17.5K queries) and a smaller expert-annotated evaluation set (392
queries). Using this data, we compare various information retrieval algorithms
on our test set and present a superior bi-encoder retriever for text-based
dataset recommendation. This system, trained on the DataFinder Dataset, finds
more relevant search results than existing third-party dataset search engines.
To encourage progress on dataset recommendation, we release our dataset and
models to the public.
| [
{
"version": "v1",
"created": "Fri, 26 May 2023 05:22:36 GMT"
},
{
"version": "v2",
"created": "Wed, 7 Jun 2023 03:08:27 GMT"
}
] | 2023-06-08T00:00:00 | [
[
"Viswanathan",
"Vijay",
""
],
[
"Gao",
"Luyu",
""
],
[
"Wu",
"Tongshuang",
""
],
[
"Liu",
"Pengfei",
""
],
[
"Neubig",
"Graham",
""
]
] | TITLE: DataFinder: Scientific Dataset Recommendation from Natural Language
Descriptions
ABSTRACT: Modern machine learning relies on datasets to develop and validate research
ideas. Given the growth of publicly available data, finding the right dataset
to use is increasingly difficult. Any research question imposes explicit and
implicit constraints on how well a given dataset will enable researchers to
answer this question, such as dataset size, modality, and domain. We
operationalize the task of recommending datasets given a short natural language
description of a research idea, to help people find relevant datasets for their
needs. Dataset recommendation poses unique challenges as an information
retrieval problem; datasets are hard to directly index for search and there are
no corpora readily available for this task. To facilitate this task, we build
the DataFinder Dataset which consists of a larger automatically-constructed
training set (17.5K queries) and a smaller expert-annotated evaluation set (392
queries). Using this data, we compare various information retrieval algorithms
on our test set and present a superior bi-encoder retriever for text-based
dataset recommendation. This system, trained on the DataFinder Dataset, finds
more relevant search results than existing third-party dataset search engines.
To encourage progress on dataset recommendation, we release our dataset and
models to the public.
|
2305.16718 | V\'it Novotn\'y | V\'it Novotn\'y, Krist\'yna Luger, Michal \v{S}tef\'anik, Tereza
Vrabcov\'a, Ale\v{s} Hor\'ak | People and Places of Historical Europe: Bootstrapping Annotation
Pipeline and a New Corpus of Named Entities in Late Medieval Texts | To appear in the Findings of the Association for Computational
Linguistics: ACL 2023 | null | null | null | cs.CL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Although pre-trained named entity recognition (NER) models are highly
accurate on modern corpora, they underperform on historical texts due to
differences in language OCR errors. In this work, we develop a new NER corpus
of 3.6M sentences from late medieval charters written mainly in Czech, Latin,
and German.
We show that we can start with a list of known historical figures and
locations and an unannotated corpus of historical texts, and use information
retrieval techniques to automatically bootstrap a NER-annotated corpus. Using
our corpus, we train a NER model that achieves entity-level Precision of
72.81-93.98% with 58.14-81.77% Recall on a manually-annotated test dataset.
Furthermore, we show that using a weighted loss function helps to combat class
imbalance in token classification tasks. To make it easy for others to
reproduce and build upon our work, we publicly release our corpus, models, and
experimental code.
| [
{
"version": "v1",
"created": "Fri, 26 May 2023 08:05:01 GMT"
},
{
"version": "v2",
"created": "Tue, 6 Jun 2023 20:42:10 GMT"
}
] | 2023-06-08T00:00:00 | [
[
"Novotný",
"Vít",
""
],
[
"Luger",
"Kristýna",
""
],
[
"Štefánik",
"Michal",
""
],
[
"Vrabcová",
"Tereza",
""
],
[
"Horák",
"Aleš",
""
]
] | TITLE: People and Places of Historical Europe: Bootstrapping Annotation
Pipeline and a New Corpus of Named Entities in Late Medieval Texts
ABSTRACT: Although pre-trained named entity recognition (NER) models are highly
accurate on modern corpora, they underperform on historical texts due to
differences in language OCR errors. In this work, we develop a new NER corpus
of 3.6M sentences from late medieval charters written mainly in Czech, Latin,
and German.
We show that we can start with a list of known historical figures and
locations and an unannotated corpus of historical texts, and use information
retrieval techniques to automatically bootstrap a NER-annotated corpus. Using
our corpus, we train a NER model that achieves entity-level Precision of
72.81-93.98% with 58.14-81.77% Recall on a manually-annotated test dataset.
Furthermore, we show that using a weighted loss function helps to combat class
imbalance in token classification tasks. To make it easy for others to
reproduce and build upon our work, we publicly release our corpus, models, and
experimental code.
|
2305.16812 | Han Liu | Han Liu, Sen Chen, Ruitao Feng, Chengwei Liu, Kaixuan Li, Zhengzi Xu,
Liming Nie, Yang Liu, Yixiang Chen | A Comprehensive Study on Quality Assurance Tools for Java | 11 pages, ISSTA-2023 | null | 10.1145/3597926.3598056 | null | cs.SE | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Quality assurance (QA) tools are receiving more and more attention and are
widely used by developers. Given the wide range of solutions for QA technology,
it is still a question of evaluating QA tools. Most existing research is
limited in the following ways: (i) They compare tools without considering
scanning rules analysis. (ii) They disagree on the effectiveness of tools due
to the study methodology and benchmark dataset. (iii) They do not separately
analyze the role of the warnings. (iv) There is no large-scale study on the
analysis of time performance. To address these problems, in the paper, we
systematically select 6 free or open-source tools for a comprehensive study
from a list of 148 existing Java QA tools. To carry out a comprehensive study
and evaluate tools in multi-level dimensions, we first mapped the scanning
rules to the CWE and analyze the coverage and granularity of the scanning
rules. Then we conducted an experiment on 5 benchmarks, including 1,425 bugs,
to investigate the effectiveness of these tools. Furthermore, we took
substantial effort to investigate the effectiveness of warnings by comparing
the real labeled bugs with the warnings and investigating their role in bug
detection. Finally, we assessed these tools' time performance on 1,049
projects. The useful findings based on our comprehensive study can help
developers improve their tools and provide users with suggestions for selecting
QA tools.
| [
{
"version": "v1",
"created": "Fri, 26 May 2023 10:48:02 GMT"
},
{
"version": "v2",
"created": "Wed, 7 Jun 2023 11:45:03 GMT"
}
] | 2023-06-08T00:00:00 | [
[
"Liu",
"Han",
""
],
[
"Chen",
"Sen",
""
],
[
"Feng",
"Ruitao",
""
],
[
"Liu",
"Chengwei",
""
],
[
"Li",
"Kaixuan",
""
],
[
"Xu",
"Zhengzi",
""
],
[
"Nie",
"Liming",
""
],
[
"Liu",
"Yang",
""
],
[
"Chen",
"Yixiang",
""
]
] | TITLE: A Comprehensive Study on Quality Assurance Tools for Java
ABSTRACT: Quality assurance (QA) tools are receiving more and more attention and are
widely used by developers. Given the wide range of solutions for QA technology,
it is still a question of evaluating QA tools. Most existing research is
limited in the following ways: (i) They compare tools without considering
scanning rules analysis. (ii) They disagree on the effectiveness of tools due
to the study methodology and benchmark dataset. (iii) They do not separately
analyze the role of the warnings. (iv) There is no large-scale study on the
analysis of time performance. To address these problems, in the paper, we
systematically select 6 free or open-source tools for a comprehensive study
from a list of 148 existing Java QA tools. To carry out a comprehensive study
and evaluate tools in multi-level dimensions, we first mapped the scanning
rules to the CWE and analyze the coverage and granularity of the scanning
rules. Then we conducted an experiment on 5 benchmarks, including 1,425 bugs,
to investigate the effectiveness of these tools. Furthermore, we took
substantial effort to investigate the effectiveness of warnings by comparing
the real labeled bugs with the warnings and investigating their role in bug
detection. Finally, we assessed these tools' time performance on 1,049
projects. The useful findings based on our comprehensive study can help
developers improve their tools and provide users with suggestions for selecting
QA tools.
|
2305.18146 | Fu-An Chao | Fu-An Chao, Tien-Hong Lo, Tzu-I Wu, Yao-Ting Sung, Berlin Chen | A Hierarchical Context-aware Modeling Approach for Multi-aspect and
Multi-granular Pronunciation Assessment | Accepted to Interspeech 2023 | null | null | null | eess.AS cs.SD eess.SP | http://creativecommons.org/licenses/by/4.0/ | Automatic Pronunciation Assessment (APA) plays a vital role in
Computer-assisted Pronunciation Training (CAPT) when evaluating a second
language (L2) learner's speaking proficiency. However, an apparent downside of
most de facto methods is that they parallelize the modeling process throughout
different speech granularities without accounting for the hierarchical and
local contextual relationships among them. In light of this, a novel
hierarchical approach is proposed in this paper for multi-aspect and
multi-granular APA. Specifically, we first introduce the notion of sup-phonemes
to explore more subtle semantic traits of L2 speakers. Second, a depth-wise
separable convolution layer is exploited to better encapsulate the local
context cues at the sub-word level. Finally, we use a score-restraint attention
pooling mechanism to predict the sentence-level scores and optimize the
component models with a multitask learning (MTL) framework. Extensive
experiments carried out on a publicly-available benchmark dataset, viz.
speechocean762, demonstrate the efficacy of our approach in relation to some
cutting-edge baselines.
| [
{
"version": "v1",
"created": "Mon, 29 May 2023 15:17:32 GMT"
},
{
"version": "v2",
"created": "Fri, 2 Jun 2023 02:08:24 GMT"
},
{
"version": "v3",
"created": "Tue, 6 Jun 2023 01:41:20 GMT"
},
{
"version": "v4",
"created": "Wed, 7 Jun 2023 15:41:18 GMT"
}
] | 2023-06-08T00:00:00 | [
[
"Chao",
"Fu-An",
""
],
[
"Lo",
"Tien-Hong",
""
],
[
"Wu",
"Tzu-I",
""
],
[
"Sung",
"Yao-Ting",
""
],
[
"Chen",
"Berlin",
""
]
] | TITLE: A Hierarchical Context-aware Modeling Approach for Multi-aspect and
Multi-granular Pronunciation Assessment
ABSTRACT: Automatic Pronunciation Assessment (APA) plays a vital role in
Computer-assisted Pronunciation Training (CAPT) when evaluating a second
language (L2) learner's speaking proficiency. However, an apparent downside of
most de facto methods is that they parallelize the modeling process throughout
different speech granularities without accounting for the hierarchical and
local contextual relationships among them. In light of this, a novel
hierarchical approach is proposed in this paper for multi-aspect and
multi-granular APA. Specifically, we first introduce the notion of sup-phonemes
to explore more subtle semantic traits of L2 speakers. Second, a depth-wise
separable convolution layer is exploited to better encapsulate the local
context cues at the sub-word level. Finally, we use a score-restraint attention
pooling mechanism to predict the sentence-level scores and optimize the
component models with a multitask learning (MTL) framework. Extensive
experiments carried out on a publicly-available benchmark dataset, viz.
speechocean762, demonstrate the efficacy of our approach in relation to some
cutting-edge baselines.
|
2306.00088 | Yuxin Tang | Yuxin Tang, Zhimin Ding, Dimitrije Jankov, Binhang Yuan, Daniel
Bourgeois, Chris Jermaine | Auto-Differentiation of Relational Computations for Very Large Scale
Machine Learning | ICML 2023 | null | null | null | cs.LG cs.DB | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The relational data model was designed to facilitate large-scale data
management and analytics. We consider the problem of how to differentiate
computations expressed relationally. We show experimentally that a relational
engine running an auto-differentiated relational algorithm can easily scale to
very large datasets, and is competitive with state-of-the-art, special-purpose
systems for large-scale distributed machine learning.
| [
{
"version": "v1",
"created": "Wed, 31 May 2023 18:06:28 GMT"
},
{
"version": "v2",
"created": "Wed, 7 Jun 2023 04:25:10 GMT"
}
] | 2023-06-08T00:00:00 | [
[
"Tang",
"Yuxin",
""
],
[
"Ding",
"Zhimin",
""
],
[
"Jankov",
"Dimitrije",
""
],
[
"Yuan",
"Binhang",
""
],
[
"Bourgeois",
"Daniel",
""
],
[
"Jermaine",
"Chris",
""
]
] | TITLE: Auto-Differentiation of Relational Computations for Very Large Scale
Machine Learning
ABSTRACT: The relational data model was designed to facilitate large-scale data
management and analytics. We consider the problem of how to differentiate
computations expressed relationally. We show experimentally that a relational
engine running an auto-differentiated relational algorithm can easily scale to
very large datasets, and is competitive with state-of-the-art, special-purpose
systems for large-scale distributed machine learning.
|
2306.00390 | Jiewen Deng | Jiewen Deng, Jinliang Deng, Renhe Jiang, Xuan Song | Learning Gaussian Mixture Representations for Tensor Time Series
Forecasting | Accepted by IJCAI 2023 Main Track | null | null | null | cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Tensor time series (TTS) data, a generalization of one-dimensional time
series on a high-dimensional space, is ubiquitous in real-world scenarios,
especially in monitoring systems involving multi-source spatio-temporal data
(e.g., transportation demands and air pollutants). Compared to modeling time
series or multivariate time series, which has received much attention and
achieved tremendous progress in recent years, tensor time series has been paid
less effort. Properly coping with the tensor time series is a much more
challenging task, due to its high-dimensional and complex inner structure. In
this paper, we develop a novel TTS forecasting framework, which seeks to
individually model each heterogeneity component implied in the time, the
location, and the source variables. We name this framework as GMRL, short for
Gaussian Mixture Representation Learning. Experiment results on two real-world
TTS datasets verify the superiority of our approach compared with the
state-of-the-art baselines. Code and data are published on
https://github.com/beginner-sketch/GMRL.
| [
{
"version": "v1",
"created": "Thu, 1 Jun 2023 06:50:47 GMT"
},
{
"version": "v2",
"created": "Tue, 6 Jun 2023 13:18:50 GMT"
},
{
"version": "v3",
"created": "Wed, 7 Jun 2023 13:07:38 GMT"
}
] | 2023-06-08T00:00:00 | [
[
"Deng",
"Jiewen",
""
],
[
"Deng",
"Jinliang",
""
],
[
"Jiang",
"Renhe",
""
],
[
"Song",
"Xuan",
""
]
] | TITLE: Learning Gaussian Mixture Representations for Tensor Time Series
Forecasting
ABSTRACT: Tensor time series (TTS) data, a generalization of one-dimensional time
series on a high-dimensional space, is ubiquitous in real-world scenarios,
especially in monitoring systems involving multi-source spatio-temporal data
(e.g., transportation demands and air pollutants). Compared to modeling time
series or multivariate time series, which has received much attention and
achieved tremendous progress in recent years, tensor time series has been paid
less effort. Properly coping with the tensor time series is a much more
challenging task, due to its high-dimensional and complex inner structure. In
this paper, we develop a novel TTS forecasting framework, which seeks to
individually model each heterogeneity component implied in the time, the
location, and the source variables. We name this framework as GMRL, short for
Gaussian Mixture Representation Learning. Experiment results on two real-world
TTS datasets verify the superiority of our approach compared with the
state-of-the-art baselines. Code and data are published on
https://github.com/beginner-sketch/GMRL.
|
2306.00823 | Sebastian Bullinger | Sebastian Bullinger and Florian Fervers and Christoph Bodensteiner and
Michael Arens | Geo-Tiles for Semantic Segmentation of Earth Observation Imagery | null | null | null | null | cs.CV eess.IV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | To cope with the high requirements during the computation of semantic
segmentations of earth observation imagery, current state-of-the-art pipelines
divide the corresponding data into smaller images. Existing methods and
benchmark datasets oftentimes rely on pixel-based tiling schemes or on
geo-tiling schemes employed by web mapping applications. The selection of
subimages (comprising size, location and orientation) is crucial. It affects
the available context information of each pixel, defines the number of tiles
during training, and influences the degree of information degradation while
down- and up-sampling the tile contents to the size required by the
segmentation model. We propose a new segmentation pipeline for earth
observation imagery relying on a tiling scheme that creates geo-tiles based on
the geo-information of the raster data. This approach exhibits several
beneficial properties compared to pixel-based or common web mapping approaches.
The proposed tiling scheme shows flexible customization properties regarding
tile granularity, tile stride and image boundary alignment. This allows us to
perform a tile specific data augmentation during training and a substitution of
pixel predictions with limited context information using data of overlapping
tiles during inference. The generated tiles show a consistent spatial tile
extent w.r.t. heterogeneous sensors, varying recording distances and different
latitudes. We demonstrate how the proposed tiling system allows to improve the
results of current state-of-the-art semantic segmentation models. To foster
future research we make the source code publicly available.
| [
{
"version": "v1",
"created": "Thu, 1 Jun 2023 15:46:04 GMT"
},
{
"version": "v2",
"created": "Wed, 7 Jun 2023 16:23:59 GMT"
}
] | 2023-06-08T00:00:00 | [
[
"Bullinger",
"Sebastian",
""
],
[
"Fervers",
"Florian",
""
],
[
"Bodensteiner",
"Christoph",
""
],
[
"Arens",
"Michael",
""
]
] | TITLE: Geo-Tiles for Semantic Segmentation of Earth Observation Imagery
ABSTRACT: To cope with the high requirements during the computation of semantic
segmentations of earth observation imagery, current state-of-the-art pipelines
divide the corresponding data into smaller images. Existing methods and
benchmark datasets oftentimes rely on pixel-based tiling schemes or on
geo-tiling schemes employed by web mapping applications. The selection of
subimages (comprising size, location and orientation) is crucial. It affects
the available context information of each pixel, defines the number of tiles
during training, and influences the degree of information degradation while
down- and up-sampling the tile contents to the size required by the
segmentation model. We propose a new segmentation pipeline for earth
observation imagery relying on a tiling scheme that creates geo-tiles based on
the geo-information of the raster data. This approach exhibits several
beneficial properties compared to pixel-based or common web mapping approaches.
The proposed tiling scheme shows flexible customization properties regarding
tile granularity, tile stride and image boundary alignment. This allows us to
perform a tile specific data augmentation during training and a substitution of
pixel predictions with limited context information using data of overlapping
tiles during inference. The generated tiles show a consistent spatial tile
extent w.r.t. heterogeneous sensors, varying recording distances and different
latitudes. We demonstrate how the proposed tiling system allows to improve the
results of current state-of-the-art semantic segmentation models. To foster
future research we make the source code publicly available.
|
2306.01562 | Jiayuan Wang | Jiayuan Wang, Q. M. Jonathan Wu and Farhad Pourpanah | An Attentive-based Generative Model for Medical Image Synthesis | null | null | 10.1007/s13042-023-01871-0 | null | eess.IV cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Magnetic resonance (MR) and computer tomography (CT) imaging are valuable
tools for diagnosing diseases and planning treatment. However, limitations such
as radiation exposure and cost can restrict access to certain imaging
modalities. To address this issue, medical image synthesis can generate one
modality from another, but many existing models struggle with high-quality
image synthesis when multiple slices are present in the dataset. This study
proposes an attention-based dual contrast generative model, called
ADC-cycleGAN, which can synthesize medical images from unpaired data with
multiple slices. The model integrates a dual contrast loss term with the
CycleGAN loss to ensure that the synthesized images are distinguishable from
the source domain. Additionally, an attention mechanism is incorporated into
the generators to extract informative features from both channel and spatial
domains. To improve performance when dealing with multiple slices, the
$K$-means algorithm is used to cluster the dataset into $K$ groups, and each
group is used to train a separate ADC-cycleGAN. Experimental results
demonstrate that the proposed ADC-cycleGAN model produces comparable samples to
other state-of-the-art generative models, achieving the highest PSNR and SSIM
values of 19.04385 and 0.68551, respectively. We publish the code at
https://github.com/JiayuanWang-JW/ADC-cycleGAN.
| [
{
"version": "v1",
"created": "Fri, 2 Jun 2023 14:17:37 GMT"
}
] | 2023-06-08T00:00:00 | [
[
"Wang",
"Jiayuan",
""
],
[
"Wu",
"Q. M. Jonathan",
""
],
[
"Pourpanah",
"Farhad",
""
]
] | TITLE: An Attentive-based Generative Model for Medical Image Synthesis
ABSTRACT: Magnetic resonance (MR) and computer tomography (CT) imaging are valuable
tools for diagnosing diseases and planning treatment. However, limitations such
as radiation exposure and cost can restrict access to certain imaging
modalities. To address this issue, medical image synthesis can generate one
modality from another, but many existing models struggle with high-quality
image synthesis when multiple slices are present in the dataset. This study
proposes an attention-based dual contrast generative model, called
ADC-cycleGAN, which can synthesize medical images from unpaired data with
multiple slices. The model integrates a dual contrast loss term with the
CycleGAN loss to ensure that the synthesized images are distinguishable from
the source domain. Additionally, an attention mechanism is incorporated into
the generators to extract informative features from both channel and spatial
domains. To improve performance when dealing with multiple slices, the
$K$-means algorithm is used to cluster the dataset into $K$ groups, and each
group is used to train a separate ADC-cycleGAN. Experimental results
demonstrate that the proposed ADC-cycleGAN model produces comparable samples to
other state-of-the-art generative models, achieving the highest PSNR and SSIM
values of 19.04385 and 0.68551, respectively. We publish the code at
https://github.com/JiayuanWang-JW/ADC-cycleGAN.
|
2306.01707 | Tianduo Wang | Tianduo Wang and Wei Lu | Learning Multi-Step Reasoning by Solving Arithmetic Tasks | ACL 2023. Code and data are available at
https://github.com/TianduoWang/MsAT | null | null | null | cs.CL | http://creativecommons.org/licenses/by/4.0/ | Mathematical reasoning is regarded as a necessary ability for Language Models
(LMs). Recent works demonstrate large LMs' impressive performance in solving
math problems. The success is attributed to their Chain-of-Thought (CoT)
reasoning abilities, i.e., the ability to decompose complex questions into
step-by-step reasoning chains, but such ability seems only to emerge from
models with abundant parameters. This work investigates how to incorporate
relatively small LMs with the capabilities of multi-step reasoning. We propose
to inject such abilities by continually pre-training LMs on a synthetic dataset
MsAT which is composed of Multi-step Arithmetic Tasks. Our experiments on four
math word problem datasets show the effectiveness of the proposed method in
enhancing LMs' math reasoning abilities.
| [
{
"version": "v1",
"created": "Fri, 2 Jun 2023 17:29:22 GMT"
},
{
"version": "v2",
"created": "Tue, 6 Jun 2023 04:08:04 GMT"
},
{
"version": "v3",
"created": "Wed, 7 Jun 2023 03:45:15 GMT"
}
] | 2023-06-08T00:00:00 | [
[
"Wang",
"Tianduo",
""
],
[
"Lu",
"Wei",
""
]
] | TITLE: Learning Multi-Step Reasoning by Solving Arithmetic Tasks
ABSTRACT: Mathematical reasoning is regarded as a necessary ability for Language Models
(LMs). Recent works demonstrate large LMs' impressive performance in solving
math problems. The success is attributed to their Chain-of-Thought (CoT)
reasoning abilities, i.e., the ability to decompose complex questions into
step-by-step reasoning chains, but such ability seems only to emerge from
models with abundant parameters. This work investigates how to incorporate
relatively small LMs with the capabilities of multi-step reasoning. We propose
to inject such abilities by continually pre-training LMs on a synthetic dataset
MsAT which is composed of Multi-step Arithmetic Tasks. Our experiments on four
math word problem datasets show the effectiveness of the proposed method in
enhancing LMs' math reasoning abilities.
|
2306.02349 | Momchil Hardalov | Momchil Hardalov, Pepa Atanasova, Todor Mihaylov, Galia Angelova,
Kiril Simov, Petya Osenova, Ves Stoyanov, Ivan Koychev, Preslav Nakov,
Dragomir Radev | bgGLUE: A Bulgarian General Language Understanding Evaluation Benchmark | Accepted to ACL 2023 (Main Conference) | ACL 2023 | null | null | cs.CL cs.IR cs.LG | http://creativecommons.org/licenses/by/4.0/ | We present bgGLUE(Bulgarian General Language Understanding Evaluation), a
benchmark for evaluating language models on Natural Language Understanding
(NLU) tasks in Bulgarian. Our benchmark includes NLU tasks targeting a variety
of NLP problems (e.g., natural language inference, fact-checking, named entity
recognition, sentiment analysis, question answering, etc.) and machine learning
tasks (sequence labeling, document-level classification, and regression). We
run the first systematic evaluation of pre-trained language models for
Bulgarian, comparing and contrasting results across the nine tasks in the
benchmark. The evaluation results show strong performance on sequence labeling
tasks, but there is a lot of room for improvement for tasks that require more
complex reasoning. We make bgGLUE publicly available together with the
fine-tuning and the evaluation code, as well as a public leaderboard at
https://bgglue.github.io/, and we hope that it will enable further advancements
in developing NLU models for Bulgarian.
| [
{
"version": "v1",
"created": "Sun, 4 Jun 2023 12:54:00 GMT"
},
{
"version": "v2",
"created": "Wed, 7 Jun 2023 03:57:51 GMT"
}
] | 2023-06-08T00:00:00 | [
[
"Hardalov",
"Momchil",
""
],
[
"Atanasova",
"Pepa",
""
],
[
"Mihaylov",
"Todor",
""
],
[
"Angelova",
"Galia",
""
],
[
"Simov",
"Kiril",
""
],
[
"Osenova",
"Petya",
""
],
[
"Stoyanov",
"Ves",
""
],
[
"Koychev",
"Ivan",
""
],
[
"Nakov",
"Preslav",
""
],
[
"Radev",
"Dragomir",
""
]
] | TITLE: bgGLUE: A Bulgarian General Language Understanding Evaluation Benchmark
ABSTRACT: We present bgGLUE(Bulgarian General Language Understanding Evaluation), a
benchmark for evaluating language models on Natural Language Understanding
(NLU) tasks in Bulgarian. Our benchmark includes NLU tasks targeting a variety
of NLP problems (e.g., natural language inference, fact-checking, named entity
recognition, sentiment analysis, question answering, etc.) and machine learning
tasks (sequence labeling, document-level classification, and regression). We
run the first systematic evaluation of pre-trained language models for
Bulgarian, comparing and contrasting results across the nine tasks in the
benchmark. The evaluation results show strong performance on sequence labeling
tasks, but there is a lot of room for improvement for tasks that require more
complex reasoning. We make bgGLUE publicly available together with the
fine-tuning and the evaluation code, as well as a public leaderboard at
https://bgglue.github.io/, and we hope that it will enable further advancements
in developing NLU models for Bulgarian.
|
2306.02606 | Zhuojia Fu Prof. | Zhuojia Fu, Wenzhi Xu, Shuainan Liu | Physics-Informed Kernel Function Neural Networks for Solving Partial
Differential Equations | null | null | null | null | math.NA cs.NA | http://creativecommons.org/licenses/by-nc-nd/4.0/ | This paper proposed a novel radial basis function neural network (RBFNN) to
solve various partial differential equations (PDEs). In the proposed RBF neural
networks, the physics-informed kernel functions (PIKFs), which are derived
according to the governing equations of the considered PDEs, are used to be the
activation functions instead of the traditional RBFs. Similar to the well-known
physics-informed neural networks (PINNs), the proposed physics-informed kernel
function neural networks (PIKFNNs) also include the physical information of the
considered PDEs in the neural network. The difference is that the PINNs put
this physical information in the loss function, and the proposed PIKFNNs put
the physical information of the considered governing equations in the
activation functions. By using the derived physics-informed kernel functions
satisfying the considered governing equations of homogeneous, nonhomogeneous,
transient PDEs as the activation functions, only the boundary/initial data are
required to train the neural network. Finally, the feasibility and accuracy of
the proposed PIKFNNs are validated by several benchmark examples referred to
high-wavenumber wave propagation problem, infinite domain problem,
nonhomogeneous problem, long-time evolution problem, inverse problem, spatial
structural derivative diffusion model, and so on.
| [
{
"version": "v1",
"created": "Mon, 5 Jun 2023 05:36:29 GMT"
},
{
"version": "v2",
"created": "Wed, 7 Jun 2023 14:36:27 GMT"
}
] | 2023-06-08T00:00:00 | [
[
"Fu",
"Zhuojia",
""
],
[
"Xu",
"Wenzhi",
""
],
[
"Liu",
"Shuainan",
""
]
] | TITLE: Physics-Informed Kernel Function Neural Networks for Solving Partial
Differential Equations
ABSTRACT: This paper proposed a novel radial basis function neural network (RBFNN) to
solve various partial differential equations (PDEs). In the proposed RBF neural
networks, the physics-informed kernel functions (PIKFs), which are derived
according to the governing equations of the considered PDEs, are used to be the
activation functions instead of the traditional RBFs. Similar to the well-known
physics-informed neural networks (PINNs), the proposed physics-informed kernel
function neural networks (PIKFNNs) also include the physical information of the
considered PDEs in the neural network. The difference is that the PINNs put
this physical information in the loss function, and the proposed PIKFNNs put
the physical information of the considered governing equations in the
activation functions. By using the derived physics-informed kernel functions
satisfying the considered governing equations of homogeneous, nonhomogeneous,
transient PDEs as the activation functions, only the boundary/initial data are
required to train the neural network. Finally, the feasibility and accuracy of
the proposed PIKFNNs are validated by several benchmark examples referred to
high-wavenumber wave propagation problem, infinite domain problem,
nonhomogeneous problem, long-time evolution problem, inverse problem, spatial
structural derivative diffusion model, and so on.
|
2306.03291 | Hyun Dong Lee | Hyun Dong Lee, Andrew Warrington, Joshua I. Glaser, Scott W. Linderman | Switching Autoregressive Low-rank Tensor Models | null | null | null | null | cs.LG stat.ME stat.ML | http://creativecommons.org/licenses/by-nc-sa/4.0/ | An important problem in time-series analysis is modeling systems with
time-varying dynamics. Probabilistic models with joint continuous and discrete
latent states offer interpretable, efficient, and experimentally useful
descriptions of such data. Commonly used models include autoregressive hidden
Markov models (ARHMMs) and switching linear dynamical systems (SLDSs), each
with its own advantages and disadvantages. ARHMMs permit exact inference and
easy parameter estimation, but are parameter intensive when modeling long
dependencies, and hence are prone to overfitting. In contrast, SLDSs can
capture long-range dependencies in a parameter efficient way through Markovian
latent dynamics, but present an intractable likelihood and a challenging
parameter estimation task. In this paper, we propose switching autoregressive
low-rank tensor (SALT) models, which retain the advantages of both approaches
while ameliorating the weaknesses. SALT parameterizes the tensor of an ARHMM
with a low-rank factorization to control the number of parameters and allow
longer range dependencies without overfitting. We prove theoretical and discuss
practical connections between SALT, linear dynamical systems, and SLDSs. We
empirically demonstrate quantitative advantages of SALT models on a range of
simulated and real prediction tasks, including behavioral and neural datasets.
Furthermore, the learned low-rank tensor provides novel insights into temporal
dependencies within each discrete state.
| [
{
"version": "v1",
"created": "Mon, 5 Jun 2023 22:25:28 GMT"
},
{
"version": "v2",
"created": "Wed, 7 Jun 2023 00:35:34 GMT"
}
] | 2023-06-08T00:00:00 | [
[
"Lee",
"Hyun Dong",
""
],
[
"Warrington",
"Andrew",
""
],
[
"Glaser",
"Joshua I.",
""
],
[
"Linderman",
"Scott W.",
""
]
] | TITLE: Switching Autoregressive Low-rank Tensor Models
ABSTRACT: An important problem in time-series analysis is modeling systems with
time-varying dynamics. Probabilistic models with joint continuous and discrete
latent states offer interpretable, efficient, and experimentally useful
descriptions of such data. Commonly used models include autoregressive hidden
Markov models (ARHMMs) and switching linear dynamical systems (SLDSs), each
with its own advantages and disadvantages. ARHMMs permit exact inference and
easy parameter estimation, but are parameter intensive when modeling long
dependencies, and hence are prone to overfitting. In contrast, SLDSs can
capture long-range dependencies in a parameter efficient way through Markovian
latent dynamics, but present an intractable likelihood and a challenging
parameter estimation task. In this paper, we propose switching autoregressive
low-rank tensor (SALT) models, which retain the advantages of both approaches
while ameliorating the weaknesses. SALT parameterizes the tensor of an ARHMM
with a low-rank factorization to control the number of parameters and allow
longer range dependencies without overfitting. We prove theoretical and discuss
practical connections between SALT, linear dynamical systems, and SLDSs. We
empirically demonstrate quantitative advantages of SALT models on a range of
simulated and real prediction tasks, including behavioral and neural datasets.
Furthermore, the learned low-rank tensor provides novel insights into temporal
dependencies within each discrete state.
|
2306.03457 | Chen Tang | Tyler Loakman, Chen Tang and Chenghua Lin | TwistList: Resources and Baselines for Tongue Twister Generation | null | ACL 2023 | null | null | cs.CL cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Previous work in phonetically-grounded language generation has mainly focused
on domains such as lyrics and poetry. In this paper, we present work on the
generation of tongue twisters - a form of language that is required to be
phonetically conditioned to maximise sound overlap, whilst maintaining semantic
consistency with an input topic, and still being grammatically correct. We
present \textbf{TwistList}, a large annotated dataset of tongue twisters,
consisting of 2.1K+ human-authored examples. We additionally present several
benchmark systems (referred to as TwisterMisters) for the proposed task of
tongue twister generation, including models that both do and do not require
training on in-domain data. We present the results of automatic and human
evaluation to demonstrate the performance of existing mainstream pre-trained
models in this task with limited (or no) task specific training and data, and
no explicit phonetic knowledge. We find that the task of tongue twister
generation is challenging for models under these conditions, yet some models
are still capable of generating acceptable examples of this language type.
| [
{
"version": "v1",
"created": "Tue, 6 Jun 2023 07:20:51 GMT"
},
{
"version": "v2",
"created": "Wed, 7 Jun 2023 05:24:25 GMT"
}
] | 2023-06-08T00:00:00 | [
[
"Loakman",
"Tyler",
""
],
[
"Tang",
"Chen",
""
],
[
"Lin",
"Chenghua",
""
]
] | TITLE: TwistList: Resources and Baselines for Tongue Twister Generation
ABSTRACT: Previous work in phonetically-grounded language generation has mainly focused
on domains such as lyrics and poetry. In this paper, we present work on the
generation of tongue twisters - a form of language that is required to be
phonetically conditioned to maximise sound overlap, whilst maintaining semantic
consistency with an input topic, and still being grammatically correct. We
present \textbf{TwistList}, a large annotated dataset of tongue twisters,
consisting of 2.1K+ human-authored examples. We additionally present several
benchmark systems (referred to as TwisterMisters) for the proposed task of
tongue twister generation, including models that both do and do not require
training on in-domain data. We present the results of automatic and human
evaluation to demonstrate the performance of existing mainstream pre-trained
models in this task with limited (or no) task specific training and data, and
no explicit phonetic knowledge. We find that the task of tongue twister
generation is challenging for models under these conditions, yet some models
are still capable of generating acceptable examples of this language type.
|
2306.03646 | Miki Okamura | Miki Okamura, Naruya Kondo, Tatsuki Fushimi, Maki Sakamoto, and Yoichi
Ochiai | Dance Generation by Sound Symbolic Words | null | null | null | null | cs.LG cs.HC cs.SD eess.AS | http://creativecommons.org/licenses/by/4.0/ | This study introduces a novel approach to generate dance motions using
onomatopoeia as input, with the aim of enhancing creativity and diversity in
dance generation. Unlike text and music, onomatopoeia conveys rhythm and
meaning through abstract word expressions without constraints on expression and
without need for specialized knowledge. We adapt the AI Choreographer framework
and employ the Sakamoto system, a feature extraction method for onomatopoeia
focusing on phonemes and syllables. Additionally, we present a new dataset of
40 onomatopoeia-dance motion pairs collected through a user survey. Our results
demonstrate that the proposed method enables more intuitive dance generation
and can create dance motions using sound-symbolic words from a variety of
languages, including those without onomatopoeia. This highlights the potential
for diverse dance creation across different languages and cultures, accessible
to a wider audience. Qualitative samples from our model can be found at:
https://sites.google.com/view/onomatopoeia-dance/home/.
| [
{
"version": "v1",
"created": "Tue, 6 Jun 2023 13:00:47 GMT"
}
] | 2023-06-08T00:00:00 | [
[
"Okamura",
"Miki",
""
],
[
"Kondo",
"Naruya",
""
],
[
"Fushimi",
"Tatsuki",
""
],
[
"Sakamoto",
"Maki",
""
],
[
"Ochiai",
"Yoichi",
""
]
] | TITLE: Dance Generation by Sound Symbolic Words
ABSTRACT: This study introduces a novel approach to generate dance motions using
onomatopoeia as input, with the aim of enhancing creativity and diversity in
dance generation. Unlike text and music, onomatopoeia conveys rhythm and
meaning through abstract word expressions without constraints on expression and
without need for specialized knowledge. We adapt the AI Choreographer framework
and employ the Sakamoto system, a feature extraction method for onomatopoeia
focusing on phonemes and syllables. Additionally, we present a new dataset of
40 onomatopoeia-dance motion pairs collected through a user survey. Our results
demonstrate that the proposed method enables more intuitive dance generation
and can create dance motions using sound-symbolic words from a variety of
languages, including those without onomatopoeia. This highlights the potential
for diverse dance creation across different languages and cultures, accessible
to a wider audience. Qualitative samples from our model can be found at:
https://sites.google.com/view/onomatopoeia-dance/home/.
|
2306.03901 | Chenxu Hu | Chenxu Hu, Jie Fu, Chenzhuang Du, Simian Luo, Junbo Zhao, Hang Zhao | ChatDB: Augmenting LLMs with Databases as Their Symbolic Memory | null | null | null | null | cs.AI cs.CL cs.DB cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Large language models (LLMs) with memory are computationally universal.
However, mainstream LLMs are not taking full advantage of memory, and the
designs are heavily influenced by biological brains. Due to their approximate
nature and proneness to the accumulation of errors, conventional neural memory
mechanisms cannot support LLMs to simulate complex reasoning. In this paper, we
seek inspiration from modern computer architectures to augment LLMs with
symbolic memory for complex multi-hop reasoning. Such a symbolic memory
framework is instantiated as an LLM and a set of SQL databases, where the LLM
generates SQL instructions to manipulate the SQL databases. We validate the
effectiveness of the proposed memory framework on a synthetic dataset requiring
complex reasoning. The project website is available at
https://chatdatabase.github.io/ .
| [
{
"version": "v1",
"created": "Tue, 6 Jun 2023 17:58:24 GMT"
},
{
"version": "v2",
"created": "Wed, 7 Jun 2023 17:22:22 GMT"
}
] | 2023-06-08T00:00:00 | [
[
"Hu",
"Chenxu",
""
],
[
"Fu",
"Jie",
""
],
[
"Du",
"Chenzhuang",
""
],
[
"Luo",
"Simian",
""
],
[
"Zhao",
"Junbo",
""
],
[
"Zhao",
"Hang",
""
]
] | TITLE: ChatDB: Augmenting LLMs with Databases as Their Symbolic Memory
ABSTRACT: Large language models (LLMs) with memory are computationally universal.
However, mainstream LLMs are not taking full advantage of memory, and the
designs are heavily influenced by biological brains. Due to their approximate
nature and proneness to the accumulation of errors, conventional neural memory
mechanisms cannot support LLMs to simulate complex reasoning. In this paper, we
seek inspiration from modern computer architectures to augment LLMs with
symbolic memory for complex multi-hop reasoning. Such a symbolic memory
framework is instantiated as an LLM and a set of SQL databases, where the LLM
generates SQL instructions to manipulate the SQL databases. We validate the
effectiveness of the proposed memory framework on a synthetic dataset requiring
complex reasoning. The project website is available at
https://chatdatabase.github.io/ .
|
2306.03932 | Zaid Khan | Zaid Khan, Vijay Kumar BG, Samuel Schulter, Xiang Yu, Yun Fu, Manmohan
Chandraker | Q: How to Specialize Large Vision-Language Models to Data-Scarce VQA
Tasks? A: Self-Train on Unlabeled Images! | CVPR 2023 | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | Finetuning a large vision language model (VLM) on a target dataset after
large scale pretraining is a dominant paradigm in visual question answering
(VQA). Datasets for specialized tasks such as knowledge-based VQA or VQA in non
natural-image domains are orders of magnitude smaller than those for
general-purpose VQA. While collecting additional labels for specialized tasks
or domains can be challenging, unlabeled images are often available. We
introduce SelTDA (Self-Taught Data Augmentation), a strategy for finetuning
large VLMs on small-scale VQA datasets. SelTDA uses the VLM and target dataset
to build a teacher model that can generate question-answer pseudolabels
directly conditioned on an image alone, allowing us to pseudolabel unlabeled
images. SelTDA then finetunes the initial VLM on the original dataset augmented
with freshly pseudolabeled images. We describe a series of experiments showing
that our self-taught data augmentation increases robustness to adversarially
searched questions, counterfactual examples and rephrasings, improves domain
generalization, and results in greater retention of numerical reasoning skills.
The proposed strategy requires no additional annotations or architectural
modifications, and is compatible with any modern encoder-decoder multimodal
transformer. Code available at https://github.com/codezakh/SelTDA.
| [
{
"version": "v1",
"created": "Tue, 6 Jun 2023 18:00:47 GMT"
}
] | 2023-06-08T00:00:00 | [
[
"Khan",
"Zaid",
""
],
[
"BG",
"Vijay Kumar",
""
],
[
"Schulter",
"Samuel",
""
],
[
"Yu",
"Xiang",
""
],
[
"Fu",
"Yun",
""
],
[
"Chandraker",
"Manmohan",
""
]
] | TITLE: Q: How to Specialize Large Vision-Language Models to Data-Scarce VQA
Tasks? A: Self-Train on Unlabeled Images!
ABSTRACT: Finetuning a large vision language model (VLM) on a target dataset after
large scale pretraining is a dominant paradigm in visual question answering
(VQA). Datasets for specialized tasks such as knowledge-based VQA or VQA in non
natural-image domains are orders of magnitude smaller than those for
general-purpose VQA. While collecting additional labels for specialized tasks
or domains can be challenging, unlabeled images are often available. We
introduce SelTDA (Self-Taught Data Augmentation), a strategy for finetuning
large VLMs on small-scale VQA datasets. SelTDA uses the VLM and target dataset
to build a teacher model that can generate question-answer pseudolabels
directly conditioned on an image alone, allowing us to pseudolabel unlabeled
images. SelTDA then finetunes the initial VLM on the original dataset augmented
with freshly pseudolabeled images. We describe a series of experiments showing
that our self-taught data augmentation increases robustness to adversarially
searched questions, counterfactual examples and rephrasings, improves domain
generalization, and results in greater retention of numerical reasoning skills.
The proposed strategy requires no additional annotations or architectural
modifications, and is compatible with any modern encoder-decoder multimodal
transformer. Code available at https://github.com/codezakh/SelTDA.
|
2306.03934 | Constantin Seibold | Constantin Seibold, Alexander Jaus, Matthias A. Fink, Moon Kim, Simon
Rei{\ss}, Ken Herrmann, Jens Kleesiek, Rainer Stiefelhagen | Accurate Fine-Grained Segmentation of Human Anatomy in Radiographs via
Volumetric Pseudo-Labeling | 28 pages, 1 table, 10 figures | null | null | null | eess.IV cs.CV cs.LG | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Purpose: Interpreting chest radiographs (CXR) remains challenging due to the
ambiguity of overlapping structures such as the lungs, heart, and bones. To
address this issue, we propose a novel method for extracting fine-grained
anatomical structures in CXR using pseudo-labeling of three-dimensional
computed tomography (CT) scans.
Methods: We created a large-scale dataset of 10,021 thoracic CTs with 157
labels and applied an ensemble of 3D anatomy segmentation models to extract
anatomical pseudo-labels. These labels were projected onto a two-dimensional
plane, similar to the CXR, allowing the training of detailed semantic
segmentation models for CXR without any manual annotation effort.
Results: Our resulting segmentation models demonstrated remarkable
performance on CXR, with a high average model-annotator agreement between two
radiologists with mIoU scores of 0.93 and 0.85 for frontal and lateral anatomy,
while inter-annotator agreement remained at 0.95 and 0.83 mIoU. Our anatomical
segmentations allowed for the accurate extraction of relevant explainable
medical features such as the cardio-thoracic-ratio.
Conclusion: Our method of volumetric pseudo-labeling paired with CT
projection offers a promising approach for detailed anatomical segmentation of
CXR with a high agreement with human annotators. This technique may have
important clinical implications, particularly in the analysis of various
thoracic pathologies.
| [
{
"version": "v1",
"created": "Tue, 6 Jun 2023 18:01:08 GMT"
}
] | 2023-06-08T00:00:00 | [
[
"Seibold",
"Constantin",
""
],
[
"Jaus",
"Alexander",
""
],
[
"Fink",
"Matthias A.",
""
],
[
"Kim",
"Moon",
""
],
[
"Reiß",
"Simon",
""
],
[
"Herrmann",
"Ken",
""
],
[
"Kleesiek",
"Jens",
""
],
[
"Stiefelhagen",
"Rainer",
""
]
] | TITLE: Accurate Fine-Grained Segmentation of Human Anatomy in Radiographs via
Volumetric Pseudo-Labeling
ABSTRACT: Purpose: Interpreting chest radiographs (CXR) remains challenging due to the
ambiguity of overlapping structures such as the lungs, heart, and bones. To
address this issue, we propose a novel method for extracting fine-grained
anatomical structures in CXR using pseudo-labeling of three-dimensional
computed tomography (CT) scans.
Methods: We created a large-scale dataset of 10,021 thoracic CTs with 157
labels and applied an ensemble of 3D anatomy segmentation models to extract
anatomical pseudo-labels. These labels were projected onto a two-dimensional
plane, similar to the CXR, allowing the training of detailed semantic
segmentation models for CXR without any manual annotation effort.
Results: Our resulting segmentation models demonstrated remarkable
performance on CXR, with a high average model-annotator agreement between two
radiologists with mIoU scores of 0.93 and 0.85 for frontal and lateral anatomy,
while inter-annotator agreement remained at 0.95 and 0.83 mIoU. Our anatomical
segmentations allowed for the accurate extraction of relevant explainable
medical features such as the cardio-thoracic-ratio.
Conclusion: Our method of volumetric pseudo-labeling paired with CT
projection offers a promising approach for detailed anatomical segmentation of
CXR with a high agreement with human annotators. This technique may have
important clinical implications, particularly in the analysis of various
thoracic pathologies.
|
2306.03954 | Adham Atyabi | Angel I. Solis, Justin Zarkovacki, John Ly and Adham Atyabi | Recognition of Handwritten Japanese Characters Using Ensemble of
Convolutional Neural Networks | null | null | null | null | cs.CV cs.AI cs.CL cs.LG | http://creativecommons.org/licenses/by/4.0/ | The Japanese writing system is complex, with three character types of
Hiragana, Katakana, and Kanji. Kanji consists of thousands of unique
characters, further adding to the complexity of character identification and
literature understanding. Being able to translate handwritten Japanese
characters into digital text is useful for data analysis, translation, learning
and cultural preservation. In this study, a machine learning approach to
analyzing and recognizing handwritten Japanese characters (Kanji) is proposed.
The study used an ensemble of three convolutional neural networks (CNNs) for
recognizing handwritten Kanji characters and utilized four datasets of MNIST,
K-MNIST, Kuzushiji-49 (K49) and the top 150 represented classes in the
Kuzushiji-Kanji (K-Kanji) dataset for its performance evaluation. The results
indicate feasibility of using proposed CNN-ensemble architecture for
recognizing handwritten characters, achieving 99.4%, 96.4%, 95.0% and 96.4%
classification accuracy on MNIST, K-MNIS, K49, and K-Kanji datasets
respectively.
| [
{
"version": "v1",
"created": "Tue, 6 Jun 2023 18:30:51 GMT"
}
] | 2023-06-08T00:00:00 | [
[
"Solis",
"Angel I.",
""
],
[
"Zarkovacki",
"Justin",
""
],
[
"Ly",
"John",
""
],
[
"Atyabi",
"Adham",
""
]
] | TITLE: Recognition of Handwritten Japanese Characters Using Ensemble of
Convolutional Neural Networks
ABSTRACT: The Japanese writing system is complex, with three character types of
Hiragana, Katakana, and Kanji. Kanji consists of thousands of unique
characters, further adding to the complexity of character identification and
literature understanding. Being able to translate handwritten Japanese
characters into digital text is useful for data analysis, translation, learning
and cultural preservation. In this study, a machine learning approach to
analyzing and recognizing handwritten Japanese characters (Kanji) is proposed.
The study used an ensemble of three convolutional neural networks (CNNs) for
recognizing handwritten Kanji characters and utilized four datasets of MNIST,
K-MNIST, Kuzushiji-49 (K49) and the top 150 represented classes in the
Kuzushiji-Kanji (K-Kanji) dataset for its performance evaluation. The results
indicate feasibility of using proposed CNN-ensemble architecture for
recognizing handwritten characters, achieving 99.4%, 96.4%, 95.0% and 96.4%
classification accuracy on MNIST, K-MNIS, K49, and K-Kanji datasets
respectively.
|
2306.03959 | Julia White | Julia White and Arushi Raghuvanshi and Yada Pruksachatkun | Leveraging Explicit Procedural Instructions for Data-Efficient Action
Prediction | null | null | null | null | cs.CL cs.IR | http://creativecommons.org/licenses/by/4.0/ | Task-oriented dialogues often require agents to enact complex, multi-step
procedures in order to meet user requests. While large language models have
found success automating these dialogues in constrained environments, their
widespread deployment is limited by the substantial quantities of task-specific
data required for training. The following paper presents a data-efficient
solution to constructing dialogue systems, leveraging explicit instructions
derived from agent guidelines, such as company policies or customer service
manuals. Our proposed Knowledge-Augmented Dialogue System (KADS) combines a
large language model with a knowledge retrieval module that pulls documents
outlining relevant procedures from a predefined set of policies, given a
user-agent interaction. To train this system, we introduce a semi-supervised
pre-training scheme that employs dialogue-document matching and action-oriented
masked language modeling with partial parameter freezing. We evaluate the
effectiveness of our approach on prominent task-oriented dialogue datasets,
Action-Based Conversations Dataset and Schema-Guided Dialogue, for two dialogue
tasks: action state tracking and workflow discovery. Our results demonstrate
that procedural knowledge augmentation improves accuracy predicting in- and
out-of-distribution actions while preserving high performance in settings with
low or sparse data.
| [
{
"version": "v1",
"created": "Tue, 6 Jun 2023 18:42:08 GMT"
}
] | 2023-06-08T00:00:00 | [
[
"White",
"Julia",
""
],
[
"Raghuvanshi",
"Arushi",
""
],
[
"Pruksachatkun",
"Yada",
""
]
] | TITLE: Leveraging Explicit Procedural Instructions for Data-Efficient Action
Prediction
ABSTRACT: Task-oriented dialogues often require agents to enact complex, multi-step
procedures in order to meet user requests. While large language models have
found success automating these dialogues in constrained environments, their
widespread deployment is limited by the substantial quantities of task-specific
data required for training. The following paper presents a data-efficient
solution to constructing dialogue systems, leveraging explicit instructions
derived from agent guidelines, such as company policies or customer service
manuals. Our proposed Knowledge-Augmented Dialogue System (KADS) combines a
large language model with a knowledge retrieval module that pulls documents
outlining relevant procedures from a predefined set of policies, given a
user-agent interaction. To train this system, we introduce a semi-supervised
pre-training scheme that employs dialogue-document matching and action-oriented
masked language modeling with partial parameter freezing. We evaluate the
effectiveness of our approach on prominent task-oriented dialogue datasets,
Action-Based Conversations Dataset and Schema-Guided Dialogue, for two dialogue
tasks: action state tracking and workflow discovery. Our results demonstrate
that procedural knowledge augmentation improves accuracy predicting in- and
out-of-distribution actions while preserving high performance in settings with
low or sparse data.
|
2306.03962 | Amartya Sanyal | Francesco Pinto, Yaxi Hu, Fanny Yang, Amartya Sanyal | PILLAR: How to make semi-private learning more effective | null | null | null | null | cs.LG cs.AI cs.CR stat.ML | http://creativecommons.org/licenses/by/4.0/ | In Semi-Supervised Semi-Private (SP) learning, the learner has access to both
public unlabelled and private labelled data. We propose a computationally
efficient algorithm that, under mild assumptions on the data, provably achieves
significantly lower private labelled sample complexity and can be efficiently
run on real-world datasets. For this purpose, we leverage the features
extracted by networks pre-trained on public (labelled or unlabelled) data,
whose distribution can significantly differ from the one on which SP learning
is performed. To validate its empirical effectiveness, we propose a wide
variety of experiments under tight privacy constraints ($\epsilon = 0.1$) and
with a focus on low-data regimes. In all of these settings, our algorithm
exhibits significantly improved performance over available baselines that use
similar amounts of public data.
| [
{
"version": "v1",
"created": "Tue, 6 Jun 2023 18:45:05 GMT"
}
] | 2023-06-08T00:00:00 | [
[
"Pinto",
"Francesco",
""
],
[
"Hu",
"Yaxi",
""
],
[
"Yang",
"Fanny",
""
],
[
"Sanyal",
"Amartya",
""
]
] | TITLE: PILLAR: How to make semi-private learning more effective
ABSTRACT: In Semi-Supervised Semi-Private (SP) learning, the learner has access to both
public unlabelled and private labelled data. We propose a computationally
efficient algorithm that, under mild assumptions on the data, provably achieves
significantly lower private labelled sample complexity and can be efficiently
run on real-world datasets. For this purpose, we leverage the features
extracted by networks pre-trained on public (labelled or unlabelled) data,
whose distribution can significantly differ from the one on which SP learning
is performed. To validate its empirical effectiveness, we propose a wide
variety of experiments under tight privacy constraints ($\epsilon = 0.1$) and
with a focus on low-data regimes. In all of these settings, our algorithm
exhibits significantly improved performance over available baselines that use
similar amounts of public data.
|
2306.03968 | Alexander Immer | Alexander Immer, Tycho F. A. van der Ouderaa, Mark van der Wilk,
Gunnar R\"atsch, Bernhard Sch\"olkopf | Stochastic Marginal Likelihood Gradients using Neural Tangent Kernels | ICML 2023 | null | null | null | stat.ML cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Selecting hyperparameters in deep learning greatly impacts its effectiveness
but requires manual effort and expertise. Recent works show that Bayesian model
selection with Laplace approximations can allow to optimize such
hyperparameters just like standard neural network parameters using gradients
and on the training data. However, estimating a single hyperparameter gradient
requires a pass through the entire dataset, limiting the scalability of such
algorithms. In this work, we overcome this issue by introducing lower bounds to
the linearized Laplace approximation of the marginal likelihood. In contrast to
previous estimators, these bounds are amenable to stochastic-gradient-based
optimization and allow to trade off estimation accuracy against computational
complexity. We derive them using the function-space form of the linearized
Laplace, which can be estimated using the neural tangent kernel.
Experimentally, we show that the estimators can significantly accelerate
gradient-based hyperparameter optimization.
| [
{
"version": "v1",
"created": "Tue, 6 Jun 2023 19:02:57 GMT"
}
] | 2023-06-08T00:00:00 | [
[
"Immer",
"Alexander",
""
],
[
"van der Ouderaa",
"Tycho F. A.",
""
],
[
"van der Wilk",
"Mark",
""
],
[
"Rätsch",
"Gunnar",
""
],
[
"Schölkopf",
"Bernhard",
""
]
] | TITLE: Stochastic Marginal Likelihood Gradients using Neural Tangent Kernels
ABSTRACT: Selecting hyperparameters in deep learning greatly impacts its effectiveness
but requires manual effort and expertise. Recent works show that Bayesian model
selection with Laplace approximations can allow to optimize such
hyperparameters just like standard neural network parameters using gradients
and on the training data. However, estimating a single hyperparameter gradient
requires a pass through the entire dataset, limiting the scalability of such
algorithms. In this work, we overcome this issue by introducing lower bounds to
the linearized Laplace approximation of the marginal likelihood. In contrast to
previous estimators, these bounds are amenable to stochastic-gradient-based
optimization and allow to trade off estimation accuracy against computational
complexity. We derive them using the function-space form of the linearized
Laplace, which can be estimated using the neural tangent kernel.
Experimentally, we show that the estimators can significantly accelerate
gradient-based hyperparameter optimization.
|
2306.03978 | A. Taha Arslan | A. Taha Arslan | B\"{u}y\"{u}k dil modellerinin T\"{u}rk\c{c}e verisetleri ile
e\u{g}itilmesi ve ince ayarlanmas\i | 7 pages | null | null | null | cs.CL | http://creativecommons.org/licenses/by-sa/4.0/ | Large language models have advanced enormously, gained vast attraction and
are having a phase of intensed research. Some of the developed models and
training datasets have been made open-accessible. Hence these may be further
fine-tuned with some techniques to obtain specialized models for specific
tasks. When it comes to Turkish language, open-access models do not provide
satisfactory coverage. This is also observed over published datasets. In this
work, we propose some ideas to mitigate this issue: creating large Turkish
datasets, training LLMs with these and fine-tuning pre-trained models with
Turkish inputs. We report our findings on Turkish-based trainings with the
problems encountered along the way. We conclude with outcomes of these
experiments and propose ideas for further works.
--
B\"uy\"uk dil modelleri inan{\i}lmaz \"ol\c{c}\"ude geli\c{s}mekte, b\"uy\"uk
ilgi toplayarak ve \"uzerlerinde yo\u{g}un ara\c{s}tirmalarin yapildi\u{g}i bir
d\"onemdedirler. Geli\c{s}tirilen modeller ve e\u{g}itimde kullanilan
verisetlerinden bazilari a\c{c}ik eri\c{s}imli olarak sunulmaktadir. B\"oylece
ince ayarlama teknikleri uygulayarak \"ozelle\c{s}mi\c{s} g\"orevler i\c{c}in
\c{c}ali\c{s}abilir modeller elde edilmektedir. T\"urk\c{c}e s\"oz konusu
oldu\u{g}unda bu modellerinin kapsayicili\u{g}i yeterli d\"uzeyde de\u{g}ildir.
Bu durum, yayimlanan verisetlerinde de g\"ozlemlenebilir. Bunu a\c{s}manin
yollari T\"urk\c{c}e i\c{c}erikli b\"uy\"uk verisetlerinin olu\c{s}turulmasi,
b\"uy\"uk dil modellerinin bunlarla e\u{g}itilmesi ve \"onceden
e\u{g}itilmi\c{s} modellerin T\"urk\c{c}e girdilerle ince ayarlanmalari
olabilir. Bu \c{c}ali\c{s}mada a\c{c}ik eri\c{s}imli dil modelleri ve
verisetleri \"uzerinde durulmakta ve T\"urk\c{c}e temelli bazi deneyler,
kar\c{s}ila\c{s}ilan sorunlar ve sonu\c{c}lar irdelenmektedir.
| [
{
"version": "v1",
"created": "Tue, 6 Jun 2023 19:31:08 GMT"
}
] | 2023-06-08T00:00:00 | [
[
"Arslan",
"A. Taha",
""
]
] | TITLE: B\"{u}y\"{u}k dil modellerinin T\"{u}rk\c{c}e verisetleri ile
e\u{g}itilmesi ve ince ayarlanmas\i
ABSTRACT: Large language models have advanced enormously, gained vast attraction and
are having a phase of intensed research. Some of the developed models and
training datasets have been made open-accessible. Hence these may be further
fine-tuned with some techniques to obtain specialized models for specific
tasks. When it comes to Turkish language, open-access models do not provide
satisfactory coverage. This is also observed over published datasets. In this
work, we propose some ideas to mitigate this issue: creating large Turkish
datasets, training LLMs with these and fine-tuning pre-trained models with
Turkish inputs. We report our findings on Turkish-based trainings with the
problems encountered along the way. We conclude with outcomes of these
experiments and propose ideas for further works.
--
B\"uy\"uk dil modelleri inan{\i}lmaz \"ol\c{c}\"ude geli\c{s}mekte, b\"uy\"uk
ilgi toplayarak ve \"uzerlerinde yo\u{g}un ara\c{s}tirmalarin yapildi\u{g}i bir
d\"onemdedirler. Geli\c{s}tirilen modeller ve e\u{g}itimde kullanilan
verisetlerinden bazilari a\c{c}ik eri\c{s}imli olarak sunulmaktadir. B\"oylece
ince ayarlama teknikleri uygulayarak \"ozelle\c{s}mi\c{s} g\"orevler i\c{c}in
\c{c}ali\c{s}abilir modeller elde edilmektedir. T\"urk\c{c}e s\"oz konusu
oldu\u{g}unda bu modellerinin kapsayicili\u{g}i yeterli d\"uzeyde de\u{g}ildir.
Bu durum, yayimlanan verisetlerinde de g\"ozlemlenebilir. Bunu a\c{s}manin
yollari T\"urk\c{c}e i\c{c}erikli b\"uy\"uk verisetlerinin olu\c{s}turulmasi,
b\"uy\"uk dil modellerinin bunlarla e\u{g}itilmesi ve \"onceden
e\u{g}itilmi\c{s} modellerin T\"urk\c{c}e girdilerle ince ayarlanmalari
olabilir. Bu \c{c}ali\c{s}mada a\c{c}ik eri\c{s}imli dil modelleri ve
verisetleri \"uzerinde durulmakta ve T\"urk\c{c}e temelli bazi deneyler,
kar\c{s}ila\c{s}ilan sorunlar ve sonu\c{c}lar irdelenmektedir.
|
2306.03993 | Armin Danesh Pazho | Christopher Neff, Armin Danesh Pazho, Hamed Tabkhi | Real-Time Online Unsupervised Domain Adaptation for Real-World Person
Re-identification | null | null | null | null | cs.CV cs.AI cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Following the popularity of Unsupervised Domain Adaptation (UDA) in person
re-identification, the recently proposed setting of Online Unsupervised Domain
Adaptation (OUDA) attempts to bridge the gap towards practical applications by
introducing a consideration of streaming data. However, this still falls short
of truly representing real-world applications. This paper defines the setting
of Real-world Real-time Online Unsupervised Domain Adaptation (R$^2$OUDA) for
Person Re-identification. The R$^2$OUDA setting sets the stage for true
real-world real-time OUDA, bringing to light four major limitations found in
real-world applications that are often neglected in current research: system
generated person images, subset distribution selection, time-based data stream
segmentation, and a segment-based time constraint. To address all aspects of
this new R$^2$OUDA setting, this paper further proposes Real-World Real-Time
Online Streaming Mutual Mean-Teaching (R$^2$MMT), a novel multi-camera system
for real-world person re-identification. Taking a popular person
re-identification dataset, R$^2$MMT was used to construct over 100 data subsets
and train more than 3000 models, exploring the breadth of the R$^2$OUDA setting
to understand the training time and accuracy trade-offs and limitations for
real-world applications. R$^2$MMT, a real-world system able to respect the
strict constraints of the proposed R$^2$OUDA setting, achieves accuracies
within 0.1% of comparable OUDA methods that cannot be applied directly to
real-world applications.
| [
{
"version": "v1",
"created": "Tue, 6 Jun 2023 20:08:54 GMT"
}
] | 2023-06-08T00:00:00 | [
[
"Neff",
"Christopher",
""
],
[
"Pazho",
"Armin Danesh",
""
],
[
"Tabkhi",
"Hamed",
""
]
] | TITLE: Real-Time Online Unsupervised Domain Adaptation for Real-World Person
Re-identification
ABSTRACT: Following the popularity of Unsupervised Domain Adaptation (UDA) in person
re-identification, the recently proposed setting of Online Unsupervised Domain
Adaptation (OUDA) attempts to bridge the gap towards practical applications by
introducing a consideration of streaming data. However, this still falls short
of truly representing real-world applications. This paper defines the setting
of Real-world Real-time Online Unsupervised Domain Adaptation (R$^2$OUDA) for
Person Re-identification. The R$^2$OUDA setting sets the stage for true
real-world real-time OUDA, bringing to light four major limitations found in
real-world applications that are often neglected in current research: system
generated person images, subset distribution selection, time-based data stream
segmentation, and a segment-based time constraint. To address all aspects of
this new R$^2$OUDA setting, this paper further proposes Real-World Real-Time
Online Streaming Mutual Mean-Teaching (R$^2$MMT), a novel multi-camera system
for real-world person re-identification. Taking a popular person
re-identification dataset, R$^2$MMT was used to construct over 100 data subsets
and train more than 3000 models, exploring the breadth of the R$^2$OUDA setting
to understand the training time and accuracy trade-offs and limitations for
real-world applications. R$^2$MMT, a real-world system able to respect the
strict constraints of the proposed R$^2$OUDA setting, achieves accuracies
within 0.1% of comparable OUDA methods that cannot be applied directly to
real-world applications.
|
2306.04000 | Mohammad Saeed Ebrahimi Saadabadi | Mohammad Saeed Ebrahimi Saadabadi, Sahar Rahimi Malakshan, Ali Zafari,
Moktari Mostofa, Nasser M. Nasrabadi | A Quality Aware Sample-to-Sample Comparison for Face Recognition | IEEE/CVF Winter Conference on Applications of Computer Vision
(WACV23) | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Currently available face datasets mainly consist of a large number of
high-quality and a small number of low-quality samples. As a result, a Face
Recognition (FR) network fails to learn the distribution of low-quality samples
since they are less frequent during training (underrepresented). Moreover,
current state-of-the-art FR training paradigms are based on the
sample-to-center comparison (i.e., Softmax-based classifier), which results in
a lack of uniformity between train and test metrics. This work integrates a
quality-aware learning process at the sample level into the classification
training paradigm (QAFace). In this regard, Softmax centers are adaptively
guided to pay more attention to low-quality samples by using a quality-aware
function. Accordingly, QAFace adds a quality-based adjustment to the updating
procedure of the Softmax-based classifier to improve the performance on the
underrepresented low-quality samples. Our method adaptively finds and assigns
more attention to the recognizable low-quality samples in the training
datasets. In addition, QAFace ignores the unrecognizable low-quality samples
using the feature magnitude as a proxy for quality. As a result, QAFace
prevents class centers from getting distracted from the optimal direction. The
proposed method is superior to the state-of-the-art algorithms in extensive
experimental results on the CFP-FP, LFW, CPLFW, CALFW, AgeDB, IJB-B, and IJB-C
datasets.
| [
{
"version": "v1",
"created": "Tue, 6 Jun 2023 20:28:04 GMT"
}
] | 2023-06-08T00:00:00 | [
[
"Saadabadi",
"Mohammad Saeed Ebrahimi",
""
],
[
"Malakshan",
"Sahar Rahimi",
""
],
[
"Zafari",
"Ali",
""
],
[
"Mostofa",
"Moktari",
""
],
[
"Nasrabadi",
"Nasser M.",
""
]
] | TITLE: A Quality Aware Sample-to-Sample Comparison for Face Recognition
ABSTRACT: Currently available face datasets mainly consist of a large number of
high-quality and a small number of low-quality samples. As a result, a Face
Recognition (FR) network fails to learn the distribution of low-quality samples
since they are less frequent during training (underrepresented). Moreover,
current state-of-the-art FR training paradigms are based on the
sample-to-center comparison (i.e., Softmax-based classifier), which results in
a lack of uniformity between train and test metrics. This work integrates a
quality-aware learning process at the sample level into the classification
training paradigm (QAFace). In this regard, Softmax centers are adaptively
guided to pay more attention to low-quality samples by using a quality-aware
function. Accordingly, QAFace adds a quality-based adjustment to the updating
procedure of the Softmax-based classifier to improve the performance on the
underrepresented low-quality samples. Our method adaptively finds and assigns
more attention to the recognizable low-quality samples in the training
datasets. In addition, QAFace ignores the unrecognizable low-quality samples
using the feature magnitude as a proxy for quality. As a result, QAFace
prevents class centers from getting distracted from the optimal direction. The
proposed method is superior to the state-of-the-art algorithms in extensive
experimental results on the CFP-FP, LFW, CPLFW, CALFW, AgeDB, IJB-B, and IJB-C
datasets.
|
2306.04004 | Vignesh Kothapalli | Vignesh Kothapalli | Randomized Schur Complement Views for Graph Contrastive Learning | ICML 2023 | null | null | null | cs.LG cs.SI | http://creativecommons.org/licenses/by/4.0/ | We introduce a randomized topological augmentor based on Schur complements
for Graph Contrastive Learning (GCL). Given a graph laplacian matrix, the
technique generates unbiased approximations of its Schur complements and treats
the corresponding graphs as augmented views. We discuss the benefits of our
approach, provide theoretical justifications and present connections with graph
diffusion. Unlike previous efforts, we study the empirical effectiveness of the
augmentor in a controlled fashion by varying the design choices for subsequent
GCL phases, such as encoding and contrasting. Extensive experiments on node and
graph classification benchmarks demonstrate that our technique consistently
outperforms pre-defined and adaptive augmentation approaches to achieve
state-of-the-art results.
| [
{
"version": "v1",
"created": "Tue, 6 Jun 2023 20:35:20 GMT"
}
] | 2023-06-08T00:00:00 | [
[
"Kothapalli",
"Vignesh",
""
]
] | TITLE: Randomized Schur Complement Views for Graph Contrastive Learning
ABSTRACT: We introduce a randomized topological augmentor based on Schur complements
for Graph Contrastive Learning (GCL). Given a graph laplacian matrix, the
technique generates unbiased approximations of its Schur complements and treats
the corresponding graphs as augmented views. We discuss the benefits of our
approach, provide theoretical justifications and present connections with graph
diffusion. Unlike previous efforts, we study the empirical effectiveness of the
augmentor in a controlled fashion by varying the design choices for subsequent
GCL phases, such as encoding and contrasting. Extensive experiments on node and
graph classification benchmarks demonstrate that our technique consistently
outperforms pre-defined and adaptive augmentation approaches to achieve
state-of-the-art results.
|
2306.04021 | Alan Wu | Alan Wu and Michael S. Ryoo | Energy-Based Models for Cross-Modal Localization using Convolutional
Transformers | ICRA 2023 | null | null | null | cs.CV cs.AI cs.LG cs.RO | http://creativecommons.org/licenses/by/4.0/ | We present a novel framework using Energy-Based Models (EBMs) for localizing
a ground vehicle mounted with a range sensor against satellite imagery in the
absence of GPS. Lidar sensors have become ubiquitous on autonomous vehicles for
describing its surrounding environment. Map priors are typically built using
the same sensor modality for localization purposes. However, these map building
endeavors using range sensors are often expensive and time-consuming.
Alternatively, we leverage the use of satellite images as map priors, which are
widely available, easily accessible, and provide comprehensive coverage. We
propose a method using convolutional transformers that performs accurate
metric-level localization in a cross-modal manner, which is challenging due to
the drastic difference in appearance between the sparse range sensor readings
and the rich satellite imagery. We train our model end-to-end and demonstrate
our approach achieving higher accuracy than the state-of-the-art on KITTI,
Pandaset, and a custom dataset.
| [
{
"version": "v1",
"created": "Tue, 6 Jun 2023 21:27:08 GMT"
}
] | 2023-06-08T00:00:00 | [
[
"Wu",
"Alan",
""
],
[
"Ryoo",
"Michael S.",
""
]
] | TITLE: Energy-Based Models for Cross-Modal Localization using Convolutional
Transformers
ABSTRACT: We present a novel framework using Energy-Based Models (EBMs) for localizing
a ground vehicle mounted with a range sensor against satellite imagery in the
absence of GPS. Lidar sensors have become ubiquitous on autonomous vehicles for
describing its surrounding environment. Map priors are typically built using
the same sensor modality for localization purposes. However, these map building
endeavors using range sensors are often expensive and time-consuming.
Alternatively, we leverage the use of satellite images as map priors, which are
widely available, easily accessible, and provide comprehensive coverage. We
propose a method using convolutional transformers that performs accurate
metric-level localization in a cross-modal manner, which is challenging due to
the drastic difference in appearance between the sparse range sensor readings
and the rich satellite imagery. We train our model end-to-end and demonstrate
our approach achieving higher accuracy than the state-of-the-art on KITTI,
Pandaset, and a custom dataset.
|
2306.04039 | Jiaqi Zhai | Jiaqi Zhai, Zhaojie Gong, Yueming Wang, Xiao Sun, Zheng Yan, Fu Li,
Xing Liu | Revisiting Neural Retrieval on Accelerators | To appear in the 29th ACM SIGKDD Conference on Knowledge Discovery
and Data Mining (KDD 2023) | null | 10.1145/3580305.3599897 | null | cs.LG cs.IR | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Retrieval finds a small number of relevant candidates from a large corpus for
information retrieval and recommendation applications. A key component of
retrieval is to model (user, item) similarity, which is commonly represented as
the dot product of two learned embeddings. This formulation permits efficient
inference, commonly known as Maximum Inner Product Search (MIPS). Despite its
popularity, dot products cannot capture complex user-item interactions, which
are multifaceted and likely high rank. We hence examine non-dot-product
retrieval settings on accelerators, and propose \textit{mixture of logits}
(MoL), which models (user, item) similarity as an adaptive composition of
elementary similarity functions. This new formulation is expressive, capable of
modeling high rank (user, item) interactions, and further generalizes to the
long tail. When combined with a hierarchical retrieval strategy,
\textit{h-indexer}, we are able to scale up MoL to 100M corpus on a single GPU
with latency comparable to MIPS baselines. On public datasets, our approach
leads to uplifts of up to 77.3\% in hit rate (HR). Experiments on a large
recommendation surface at Meta showed strong metric gains and reduced
popularity bias, validating the proposed approach's performance and improved
generalization.
| [
{
"version": "v1",
"created": "Tue, 6 Jun 2023 22:08:42 GMT"
}
] | 2023-06-08T00:00:00 | [
[
"Zhai",
"Jiaqi",
""
],
[
"Gong",
"Zhaojie",
""
],
[
"Wang",
"Yueming",
""
],
[
"Sun",
"Xiao",
""
],
[
"Yan",
"Zheng",
""
],
[
"Li",
"Fu",
""
],
[
"Liu",
"Xing",
""
]
] | TITLE: Revisiting Neural Retrieval on Accelerators
ABSTRACT: Retrieval finds a small number of relevant candidates from a large corpus for
information retrieval and recommendation applications. A key component of
retrieval is to model (user, item) similarity, which is commonly represented as
the dot product of two learned embeddings. This formulation permits efficient
inference, commonly known as Maximum Inner Product Search (MIPS). Despite its
popularity, dot products cannot capture complex user-item interactions, which
are multifaceted and likely high rank. We hence examine non-dot-product
retrieval settings on accelerators, and propose \textit{mixture of logits}
(MoL), which models (user, item) similarity as an adaptive composition of
elementary similarity functions. This new formulation is expressive, capable of
modeling high rank (user, item) interactions, and further generalizes to the
long tail. When combined with a hierarchical retrieval strategy,
\textit{h-indexer}, we are able to scale up MoL to 100M corpus on a single GPU
with latency comparable to MIPS baselines. On public datasets, our approach
leads to uplifts of up to 77.3\% in hit rate (HR). Experiments on a large
recommendation surface at Meta showed strong metric gains and reduced
popularity bias, validating the proposed approach's performance and improved
generalization.
|
2306.04040 | Viktor Valadi | Viktor Valadi, Xinchi Qiu, Pedro Porto Buarque de Gusm\~ao, Nicholas
D. Lane, Mina Alibeigi | FedVal: Different good or different bad in federated learning | To appear in the proceedings of the USENIX Security Symposium 2023 | null | null | null | cs.LG cs.AI cs.CR | http://creativecommons.org/licenses/by/4.0/ | Federated learning (FL) systems are susceptible to attacks from malicious
actors who might attempt to corrupt the training model through various
poisoning attacks. FL also poses new challenges in addressing group bias, such
as ensuring fair performance for different demographic groups. Traditional
methods used to address such biases require centralized access to the data,
which FL systems do not have. In this paper, we present a novel approach FedVal
for both robustness and fairness that does not require any additional
information from clients that could raise privacy concerns and consequently
compromise the integrity of the FL system. To this end, we propose an
innovative score function based on a server-side validation method that
assesses client updates and determines the optimal aggregation balance between
locally-trained models. Our research shows that this approach not only provides
solid protection against poisoning attacks but can also be used to reduce group
bias and subsequently promote fairness while maintaining the system's
capability for differential privacy. Extensive experiments on the CIFAR-10,
FEMNIST, and PUMS ACSIncome datasets in different configurations demonstrate
the effectiveness of our method, resulting in state-of-the-art performances. We
have proven robustness in situations where 80% of participating clients are
malicious. Additionally, we have shown a significant increase in accuracy for
underrepresented labels from 32% to 53%, and increase in recall rate for
underrepresented features from 19% to 50%.
| [
{
"version": "v1",
"created": "Tue, 6 Jun 2023 22:11:13 GMT"
}
] | 2023-06-08T00:00:00 | [
[
"Valadi",
"Viktor",
""
],
[
"Qiu",
"Xinchi",
""
],
[
"de Gusmão",
"Pedro Porto Buarque",
""
],
[
"Lane",
"Nicholas D.",
""
],
[
"Alibeigi",
"Mina",
""
]
] | TITLE: FedVal: Different good or different bad in federated learning
ABSTRACT: Federated learning (FL) systems are susceptible to attacks from malicious
actors who might attempt to corrupt the training model through various
poisoning attacks. FL also poses new challenges in addressing group bias, such
as ensuring fair performance for different demographic groups. Traditional
methods used to address such biases require centralized access to the data,
which FL systems do not have. In this paper, we present a novel approach FedVal
for both robustness and fairness that does not require any additional
information from clients that could raise privacy concerns and consequently
compromise the integrity of the FL system. To this end, we propose an
innovative score function based on a server-side validation method that
assesses client updates and determines the optimal aggregation balance between
locally-trained models. Our research shows that this approach not only provides
solid protection against poisoning attacks but can also be used to reduce group
bias and subsequently promote fairness while maintaining the system's
capability for differential privacy. Extensive experiments on the CIFAR-10,
FEMNIST, and PUMS ACSIncome datasets in different configurations demonstrate
the effectiveness of our method, resulting in state-of-the-art performances. We
have proven robustness in situations where 80% of participating clients are
malicious. Additionally, we have shown a significant increase in accuracy for
underrepresented labels from 32% to 53%, and increase in recall rate for
underrepresented features from 19% to 50%.
|
2306.04059 | Muskan Garg | Chandreen Liyanage, Muskan Garg, Vijay Mago, Sunghwan Sohn | Augmenting Reddit Posts to Determine Wellness Dimensions impacting
Mental Health | null | null | null | null | cs.CL cs.CY | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Amid ongoing health crisis, there is a growing necessity to discern possible
signs of Wellness Dimensions (WD) manifested in self-narrated text. As the
distribution of WD on social media data is intrinsically imbalanced, we
experiment the generative NLP models for data augmentation to enable further
improvement in the pre-screening task of classifying WD. To this end, we
propose a simple yet effective data augmentation approach through prompt-based
Generative NLP models, and evaluate the ROUGE scores and syntactic/semantic
similarity among existing interpretations and augmented data. Our approach with
ChatGPT model surpasses all the other methods and achieves improvement over
baselines such as Easy-Data Augmentation and Backtranslation. Introducing data
augmentation to generate more training samples and balanced dataset, results in
the improved F-score and the Matthew's Correlation Coefficient for upto 13.11%
and 15.95%, respectively.
| [
{
"version": "v1",
"created": "Tue, 6 Jun 2023 23:15:59 GMT"
}
] | 2023-06-08T00:00:00 | [
[
"Liyanage",
"Chandreen",
""
],
[
"Garg",
"Muskan",
""
],
[
"Mago",
"Vijay",
""
],
[
"Sohn",
"Sunghwan",
""
]
] | TITLE: Augmenting Reddit Posts to Determine Wellness Dimensions impacting
Mental Health
ABSTRACT: Amid ongoing health crisis, there is a growing necessity to discern possible
signs of Wellness Dimensions (WD) manifested in self-narrated text. As the
distribution of WD on social media data is intrinsically imbalanced, we
experiment the generative NLP models for data augmentation to enable further
improvement in the pre-screening task of classifying WD. To this end, we
propose a simple yet effective data augmentation approach through prompt-based
Generative NLP models, and evaluate the ROUGE scores and syntactic/semantic
similarity among existing interpretations and augmented data. Our approach with
ChatGPT model surpasses all the other methods and achieves improvement over
baselines such as Easy-Data Augmentation and Backtranslation. Introducing data
augmentation to generate more training samples and balanced dataset, results in
the improved F-score and the Matthew's Correlation Coefficient for upto 13.11%
and 15.95%, respectively.
|
2306.04085 | Yusen Zhang | Yusen Zhang, Jun Wang, Zhiguo Wang, Rui Zhang | XSemPLR: Cross-Lingual Semantic Parsing in Multiple Natural Languages
and Meaning Representations | ACL 2023 | null | null | null | cs.CL | http://creativecommons.org/licenses/by/4.0/ | Cross-Lingual Semantic Parsing (CLSP) aims to translate queries in multiple
natural languages (NLs) into meaning representations (MRs) such as SQL, lambda
calculus, and logic forms. However, existing CLSP models are separately
proposed and evaluated on datasets of limited tasks and applications, impeding
a comprehensive and unified evaluation of CLSP on a diverse range of NLs and
MRs. To this end, we present XSemPLR, a unified benchmark for cross-lingual
semantic parsing featured with 22 natural languages and 8 meaning
representations by examining and selecting 9 existing datasets to cover 5 tasks
and 164 domains. We use XSemPLR to conduct a comprehensive benchmark study on a
wide range of multilingual language models including encoder-based models
(mBERT, XLM-R), encoder-decoder models (mBART, mT5), and decoder-based models
(Codex, BLOOM). We design 6 experiment settings covering various lingual
combinations (monolingual, multilingual, cross-lingual) and numbers of learning
samples (full dataset, few-shot, and zero-shot). Our experiments show that
encoder-decoder models (mT5) achieve the highest performance compared with
other popular models, and multilingual training can further improve the average
performance. Notably, multilingual large language models (e.g., BLOOM) are
still inadequate to perform CLSP tasks. We also find that the performance gap
between monolingual training and cross-lingual transfer learning is still
significant for multilingual models, though it can be mitigated by
cross-lingual few-shot training. Our dataset and code are available at
https://github.com/psunlpgroup/XSemPLR.
| [
{
"version": "v1",
"created": "Wed, 7 Jun 2023 01:09:37 GMT"
}
] | 2023-06-08T00:00:00 | [
[
"Zhang",
"Yusen",
""
],
[
"Wang",
"Jun",
""
],
[
"Wang",
"Zhiguo",
""
],
[
"Zhang",
"Rui",
""
]
] | TITLE: XSemPLR: Cross-Lingual Semantic Parsing in Multiple Natural Languages
and Meaning Representations
ABSTRACT: Cross-Lingual Semantic Parsing (CLSP) aims to translate queries in multiple
natural languages (NLs) into meaning representations (MRs) such as SQL, lambda
calculus, and logic forms. However, existing CLSP models are separately
proposed and evaluated on datasets of limited tasks and applications, impeding
a comprehensive and unified evaluation of CLSP on a diverse range of NLs and
MRs. To this end, we present XSemPLR, a unified benchmark for cross-lingual
semantic parsing featured with 22 natural languages and 8 meaning
representations by examining and selecting 9 existing datasets to cover 5 tasks
and 164 domains. We use XSemPLR to conduct a comprehensive benchmark study on a
wide range of multilingual language models including encoder-based models
(mBERT, XLM-R), encoder-decoder models (mBART, mT5), and decoder-based models
(Codex, BLOOM). We design 6 experiment settings covering various lingual
combinations (monolingual, multilingual, cross-lingual) and numbers of learning
samples (full dataset, few-shot, and zero-shot). Our experiments show that
encoder-decoder models (mT5) achieve the highest performance compared with
other popular models, and multilingual training can further improve the average
performance. Notably, multilingual large language models (e.g., BLOOM) are
still inadequate to perform CLSP tasks. We also find that the performance gap
between monolingual training and cross-lingual transfer learning is still
significant for multilingual models, though it can be mitigated by
cross-lingual few-shot training. Our dataset and code are available at
https://github.com/psunlpgroup/XSemPLR.
|
2306.04098 | Fiona Victoria Stanley Jothiraj | Fiona Victoria Stanley Jothiraj and Afra Mashhadi | Phoenix: A Federated Generative Diffusion Model | null | null | null | null | cs.LG cs.CV | http://creativecommons.org/licenses/by/4.0/ | Generative AI has made impressive strides in enabling users to create diverse
and realistic visual content such as images, videos, and audio. However,
training generative models on large centralized datasets can pose challenges in
terms of data privacy, security, and accessibility. Federated learning (FL) is
an approach that uses decentralized techniques to collaboratively train a
shared deep learning model while retaining the training data on individual edge
devices to preserve data privacy. This paper proposes a novel method for
training a Denoising Diffusion Probabilistic Model (DDPM) across multiple data
sources using FL techniques. Diffusion models, a newly emerging generative
model, show promising results in achieving superior quality images than
Generative Adversarial Networks (GANs). Our proposed method Phoenix is an
unconditional diffusion model that leverages strategies to improve the data
diversity of generated samples even when trained on data with statistical
heterogeneity or Non-IID (Non-Independent and Identically Distributed) data. We
demonstrate how our approach outperforms the default diffusion model in an FL
setting. These results indicate that high-quality samples can be generated by
maintaining data diversity, preserving privacy, and reducing communication
between data sources, offering exciting new possibilities in the field of
generative AI.
| [
{
"version": "v1",
"created": "Wed, 7 Jun 2023 01:43:09 GMT"
}
] | 2023-06-08T00:00:00 | [
[
"Jothiraj",
"Fiona Victoria Stanley",
""
],
[
"Mashhadi",
"Afra",
""
]
] | TITLE: Phoenix: A Federated Generative Diffusion Model
ABSTRACT: Generative AI has made impressive strides in enabling users to create diverse
and realistic visual content such as images, videos, and audio. However,
training generative models on large centralized datasets can pose challenges in
terms of data privacy, security, and accessibility. Federated learning (FL) is
an approach that uses decentralized techniques to collaboratively train a
shared deep learning model while retaining the training data on individual edge
devices to preserve data privacy. This paper proposes a novel method for
training a Denoising Diffusion Probabilistic Model (DDPM) across multiple data
sources using FL techniques. Diffusion models, a newly emerging generative
model, show promising results in achieving superior quality images than
Generative Adversarial Networks (GANs). Our proposed method Phoenix is an
unconditional diffusion model that leverages strategies to improve the data
diversity of generated samples even when trained on data with statistical
heterogeneity or Non-IID (Non-Independent and Identically Distributed) data. We
demonstrate how our approach outperforms the default diffusion model in an FL
setting. These results indicate that high-quality samples can be generated by
maintaining data diversity, preserving privacy, and reducing communication
between data sources, offering exciting new possibilities in the field of
generative AI.
|
2306.04099 | Ziting Wen | Ziting Wen, Oscar Pizarro, Stefan Williams | NTKCPL: Active Learning on Top of Self-Supervised Model by Estimating
True Coverage | null | null | null | null | cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | High annotation cost for training machine learning classifiers has driven
extensive research in active learning and self-supervised learning. Recent
research has shown that in the context of supervised learning different active
learning strategies need to be applied at various stages of the training
process to ensure improved performance over the random baseline. We refer to
the point where the number of available annotations changes the suitable active
learning strategy as the phase transition point. In this paper, we establish
that when combining active learning with self-supervised models to achieve
improved performance, the phase transition point occurs earlier. It becomes
challenging to determine which strategy should be used for previously unseen
datasets. We argue that existing active learning algorithms are heavily
influenced by the phase transition because the empirical risk over the entire
active learning pool estimated by these algorithms is inaccurate and influenced
by the number of labeled samples. To address this issue, we propose a novel
active learning strategy, neural tangent kernel clustering-pseudo-labels
(NTKCPL). It estimates empirical risk based on pseudo-labels and the model
prediction with NTK approximation. We analyze the factors affecting this
approximation error and design a pseudo-label clustering generation method to
reduce the approximation error. We validate our method on five datasets,
empirically demonstrating that it outperforms the baseline methods in most
cases and is valid over a wider range of training budgets.
| [
{
"version": "v1",
"created": "Wed, 7 Jun 2023 01:43:47 GMT"
}
] | 2023-06-08T00:00:00 | [
[
"Wen",
"Ziting",
""
],
[
"Pizarro",
"Oscar",
""
],
[
"Williams",
"Stefan",
""
]
] | TITLE: NTKCPL: Active Learning on Top of Self-Supervised Model by Estimating
True Coverage
ABSTRACT: High annotation cost for training machine learning classifiers has driven
extensive research in active learning and self-supervised learning. Recent
research has shown that in the context of supervised learning different active
learning strategies need to be applied at various stages of the training
process to ensure improved performance over the random baseline. We refer to
the point where the number of available annotations changes the suitable active
learning strategy as the phase transition point. In this paper, we establish
that when combining active learning with self-supervised models to achieve
improved performance, the phase transition point occurs earlier. It becomes
challenging to determine which strategy should be used for previously unseen
datasets. We argue that existing active learning algorithms are heavily
influenced by the phase transition because the empirical risk over the entire
active learning pool estimated by these algorithms is inaccurate and influenced
by the number of labeled samples. To address this issue, we propose a novel
active learning strategy, neural tangent kernel clustering-pseudo-labels
(NTKCPL). It estimates empirical risk based on pseudo-labels and the model
prediction with NTK approximation. We analyze the factors affecting this
approximation error and design a pseudo-label clustering generation method to
reduce the approximation error. We validate our method on five datasets,
empirically demonstrating that it outperforms the baseline methods in most
cases and is valid over a wider range of training budgets.
|
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.