id
large_stringlengths 9
16
| submitter
large_stringlengths 3
64
⌀ | authors
large_stringlengths 5
20.2k
| title
large_stringlengths 5
245
| comments
large_stringlengths 1
482
⌀ | journal-ref
large_stringlengths 4
382
⌀ | doi
large_stringlengths 9
151
⌀ | report-no
large_stringlengths 2
509
⌀ | categories
large_stringlengths 5
78
| license
large_stringclasses 9
values | abstract
large_stringlengths 83
3.55k
| versions
large listlengths 1
21
| update_date
timestamp[ms]date 2007-05-23 00:00:00
2025-05-30 00:00:00
| authors_parsed
large listlengths 1
1.35k
| prompt
large_stringlengths 166
3.64k
| label
stringclasses 2
values | prob
float64 0.5
0.73
|
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
2304.11404 | Gong Chen | Gong Chen, Yanan Zhao, Yi Wang, Kim-Hui Yap | SSN: Stockwell Scattering Network for SAR Image Change Detection | 5 pages, 6 figures | IEEE Geoscience and Remote Sensing Letters, 2023 | 10.1109/LGRS.2023.3234972 | null | cs.CV eess.IV eess.SP | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Recently, synthetic aperture radar (SAR) image change detection has become an
interesting yet challenging direction due to the presence of speckle noise.
Although both traditional and modern learning-driven methods attempted to
overcome this challenge, deep convolutional neural networks (DCNNs)-based
methods are still hindered by the lack of interpretability and the requirement
of large computation power. To overcome this drawback, wavelet scattering
network (WSN) and Fourier scattering network (FSN) are proposed. Combining
respective merits of WSN and FSN, we propose Stockwell scattering network (SSN)
based on Stockwell transform which is widely applied against noisy signals and
shows advantageous characteristics in speckle reduction. The proposed SSN
provides noise-resilient feature representation and obtains state-of-art
performance in SAR image change detection as well as high computational
efficiency. Experimental results on three real SAR image datasets demonstrate
the effectiveness of the proposed method.
| [
{
"version": "v1",
"created": "Sat, 22 Apr 2023 13:35:34 GMT"
}
] | 2023-04-25T00:00:00 | [
[
"Chen",
"Gong",
""
],
[
"Zhao",
"Yanan",
""
],
[
"Wang",
"Yi",
""
],
[
"Yap",
"Kim-Hui",
""
]
] | TITLE: SSN: Stockwell Scattering Network for SAR Image Change Detection
ABSTRACT: Recently, synthetic aperture radar (SAR) image change detection has become an
interesting yet challenging direction due to the presence of speckle noise.
Although both traditional and modern learning-driven methods attempted to
overcome this challenge, deep convolutional neural networks (DCNNs)-based
methods are still hindered by the lack of interpretability and the requirement
of large computation power. To overcome this drawback, wavelet scattering
network (WSN) and Fourier scattering network (FSN) are proposed. Combining
respective merits of WSN and FSN, we propose Stockwell scattering network (SSN)
based on Stockwell transform which is widely applied against noisy signals and
shows advantageous characteristics in speckle reduction. The proposed SSN
provides noise-resilient feature representation and obtains state-of-art
performance in SAR image change detection as well as high computational
efficiency. Experimental results on three real SAR image datasets demonstrate
the effectiveness of the proposed method.
| no_new_dataset | 0.710879 |
2304.11408 | Siddique Latif | Ahlam Husni Abu Nada, Siddique Latif, and Junaid Qadir | Lightweight Toxicity Detection in Spoken Language: A Transformer-based
Approach for Edge Devices | Under Rewiew | null | null | null | cs.SD eess.AS | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Toxicity is a prevalent social behavior that involves the use of hate speech,
offensive language, bullying, and abusive speech. While text-based approaches
for toxicity detection are common, there is limited research on processing
speech signals in the physical world. Detecting toxicity in the physical world
is challenging due to the difficulty of integrating AI-capable computers into
the environment. We propose a lightweight transformer model based on wav2vec2.0
and optimize it using techniques such as quantization and knowledge
distillation. Our model uses multitask learning and achieves an average macro
F1-score of 90.3\% and a weighted accuracy of 88\%, outperforming
state-of-the-art methods on DeToxy-B and a public dataset. Our results show
that quantization reduces the model size by almost 4 times and RAM usage by
3.3\%, with only a 1\% F1 score decrease. Knowledge distillation reduces the
model size by 3.7 times, RAM usage by 1.9, and inference time by 2 times, but
decreases accuracy by 8\%. Combining both techniques reduces the model size by
14.6 times and RAM usage by around 4.3 times, with a two-fold inference time
improvement. Our compact model is the first end-to-end speech-based toxicity
detection model based on a lightweight transformer model suitable for
deployment in physical spaces. The results show its feasibility for toxicity
detection on edge devices in real-world environments.
| [
{
"version": "v1",
"created": "Sat, 22 Apr 2023 13:45:38 GMT"
}
] | 2023-04-25T00:00:00 | [
[
"Nada",
"Ahlam Husni Abu",
""
],
[
"Latif",
"Siddique",
""
],
[
"Qadir",
"Junaid",
""
]
] | TITLE: Lightweight Toxicity Detection in Spoken Language: A Transformer-based
Approach for Edge Devices
ABSTRACT: Toxicity is a prevalent social behavior that involves the use of hate speech,
offensive language, bullying, and abusive speech. While text-based approaches
for toxicity detection are common, there is limited research on processing
speech signals in the physical world. Detecting toxicity in the physical world
is challenging due to the difficulty of integrating AI-capable computers into
the environment. We propose a lightweight transformer model based on wav2vec2.0
and optimize it using techniques such as quantization and knowledge
distillation. Our model uses multitask learning and achieves an average macro
F1-score of 90.3\% and a weighted accuracy of 88\%, outperforming
state-of-the-art methods on DeToxy-B and a public dataset. Our results show
that quantization reduces the model size by almost 4 times and RAM usage by
3.3\%, with only a 1\% F1 score decrease. Knowledge distillation reduces the
model size by 3.7 times, RAM usage by 1.9, and inference time by 2 times, but
decreases accuracy by 8\%. Combining both techniques reduces the model size by
14.6 times and RAM usage by around 4.3 times, with a two-fold inference time
improvement. Our compact model is the first end-to-end speech-based toxicity
detection model based on a lightweight transformer model suitable for
deployment in physical spaces. The results show its feasibility for toxicity
detection on edge devices in real-world environments.
| no_new_dataset | 0.712451 |
2304.11422 | Xiaowen Ma | Xiaowen Ma, Jiawei Yang, Tingfeng Hong, Mengting Ma, Ziyan Zhao, Tian
Feng and Wei Zhang | STNet: Spatial and Temporal feature fusion network for change detection
in remote sensing images | Accepted by ICME 2023 | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | As an important task in remote sensing image analysis, remote sensing change
detection (RSCD) aims to identify changes of interest in a region from
spatially co-registered multi-temporal remote sensing images, so as to monitor
the local development. Existing RSCD methods usually formulate RSCD as a binary
classification task, representing changes of interest by merely feature
concatenation or feature subtraction and recovering the spatial details via
densely connected change representations, whose performances need further
improvement. In this paper, we propose STNet, a RSCD network based on spatial
and temporal feature fusions. Specifically, we design a temporal feature fusion
(TFF) module to combine bi-temporal features using a cross-temporal gating
mechanism for emphasizing changes of interest; a spatial feature fusion module
is deployed to capture fine-grained information using a cross-scale attention
mechanism for recovering the spatial details of change representations.
Experimental results on three benchmark datasets for RSCD demonstrate that the
proposed method achieves the state-of-the-art performance. Code is available at
https://github.com/xwmaxwma/rschange.
| [
{
"version": "v1",
"created": "Sat, 22 Apr 2023 14:40:41 GMT"
}
] | 2023-04-25T00:00:00 | [
[
"Ma",
"Xiaowen",
""
],
[
"Yang",
"Jiawei",
""
],
[
"Hong",
"Tingfeng",
""
],
[
"Ma",
"Mengting",
""
],
[
"Zhao",
"Ziyan",
""
],
[
"Feng",
"Tian",
""
],
[
"Zhang",
"Wei",
""
]
] | TITLE: STNet: Spatial and Temporal feature fusion network for change detection
in remote sensing images
ABSTRACT: As an important task in remote sensing image analysis, remote sensing change
detection (RSCD) aims to identify changes of interest in a region from
spatially co-registered multi-temporal remote sensing images, so as to monitor
the local development. Existing RSCD methods usually formulate RSCD as a binary
classification task, representing changes of interest by merely feature
concatenation or feature subtraction and recovering the spatial details via
densely connected change representations, whose performances need further
improvement. In this paper, we propose STNet, a RSCD network based on spatial
and temporal feature fusions. Specifically, we design a temporal feature fusion
(TFF) module to combine bi-temporal features using a cross-temporal gating
mechanism for emphasizing changes of interest; a spatial feature fusion module
is deployed to capture fine-grained information using a cross-scale attention
mechanism for recovering the spatial details of change representations.
Experimental results on three benchmark datasets for RSCD demonstrate that the
proposed method achieves the state-of-the-art performance. Code is available at
https://github.com/xwmaxwma/rschange.
| no_new_dataset | 0.710622 |
2304.11424 | Xiaowen Ma | Xiaowen Ma, Rui Che, Tingfeng Hong, Mengting Ma, Ziyan Zhao, Tian Feng
and Wei Zhang | SACANet: scene-aware class attention network for semantic segmentation
of remote sensing images | Accepted by ICME 2023 | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Spatial attention mechanism has been widely used in semantic segmentation of
remote sensing images given its capability to model long-range dependencies.
Many methods adopting spatial attention mechanism aggregate contextual
information using direct relationships between pixels within an image, while
ignoring the scene awareness of pixels (i.e., being aware of the global context
of the scene where the pixels are located and perceiving their relative
positions). Given the observation that scene awareness benefits context
modeling with spatial correlations of ground objects, we design a scene-aware
attention module based on a refined spatial attention mechanism embedding scene
awareness. Besides, we present a local-global class attention mechanism to
address the problem that general attention mechanism introduces excessive
background noises while hardly considering the large intra-class variance in
remote sensing images. In this paper, we integrate both scene-aware and class
attentions to propose a scene-aware class attention network (SACANet) for
semantic segmentation of remote sensing images. Experimental results on three
datasets show that SACANet outperforms other state-of-the-art methods and
validate its effectiveness. Code is available at
https://github.com/xwmaxwma/rssegmentation.
| [
{
"version": "v1",
"created": "Sat, 22 Apr 2023 14:54:31 GMT"
}
] | 2023-04-25T00:00:00 | [
[
"Ma",
"Xiaowen",
""
],
[
"Che",
"Rui",
""
],
[
"Hong",
"Tingfeng",
""
],
[
"Ma",
"Mengting",
""
],
[
"Zhao",
"Ziyan",
""
],
[
"Feng",
"Tian",
""
],
[
"Zhang",
"Wei",
""
]
] | TITLE: SACANet: scene-aware class attention network for semantic segmentation
of remote sensing images
ABSTRACT: Spatial attention mechanism has been widely used in semantic segmentation of
remote sensing images given its capability to model long-range dependencies.
Many methods adopting spatial attention mechanism aggregate contextual
information using direct relationships between pixels within an image, while
ignoring the scene awareness of pixels (i.e., being aware of the global context
of the scene where the pixels are located and perceiving their relative
positions). Given the observation that scene awareness benefits context
modeling with spatial correlations of ground objects, we design a scene-aware
attention module based on a refined spatial attention mechanism embedding scene
awareness. Besides, we present a local-global class attention mechanism to
address the problem that general attention mechanism introduces excessive
background noises while hardly considering the large intra-class variance in
remote sensing images. In this paper, we integrate both scene-aware and class
attentions to propose a scene-aware class attention network (SACANet) for
semantic segmentation of remote sensing images. Experimental results on three
datasets show that SACANet outperforms other state-of-the-art methods and
validate its effectiveness. Code is available at
https://github.com/xwmaxwma/rssegmentation.
| no_new_dataset | 0.711387 |
2304.11427 | Matthew Zahr | Tianci Huang, Charles Naudet, Matthew J. Zahr | High-order implicit shock tracking boundary conditions for flows with
parametrized shocks | 29 pages, 19 figures, 1 table | null | null | null | math.NA cs.NA math.OC | http://creativecommons.org/licenses/by/4.0/ | High-order implicit shock tracking (fitting) is a class of high-order,
optimization-based numerical methods to approximate solutions of conservation
laws with non-smooth features by aligning elements of the computational mesh
with non-smooth features. This ensures the non-smooth features are perfectly
represented by inter-element jumps and high-order basis functions approximate
smooth regions of the solution without nonlinear stabilization, which leads to
accurate approximations on traditionally coarse meshes. In this work, we
introduce a robust implicit shock tracking framework specialized for problems
with parameter-dependent lead shocks (i.e., shocks separating a farfield
condition from the downstream flow), which commonly arise in high-speed
aerodynamics and astrophysics applications. After a shock-aligned mesh is
produced at one parameter configuration, all elements upstream of the lead
shock are removed and the nodes on the lead shock are positioned for new
parameter configurations using the implicit shock tracking solver. The proposed
framework can be used for most many-query applications involving parametrized
lead shocks such as optimization, uncertainty quantification, parameter sweeps,
"what-if" scenarios, or parameter-based continuation. We demonstrate the
robustness and flexibility of the framework using a one-dimensional space-time
Riemann problem, and two- and three-dimensional supersonic and hypersonic
benchmark problems.
| [
{
"version": "v1",
"created": "Sat, 22 Apr 2023 15:13:03 GMT"
}
] | 2023-04-25T00:00:00 | [
[
"Huang",
"Tianci",
""
],
[
"Naudet",
"Charles",
""
],
[
"Zahr",
"Matthew J.",
""
]
] | TITLE: High-order implicit shock tracking boundary conditions for flows with
parametrized shocks
ABSTRACT: High-order implicit shock tracking (fitting) is a class of high-order,
optimization-based numerical methods to approximate solutions of conservation
laws with non-smooth features by aligning elements of the computational mesh
with non-smooth features. This ensures the non-smooth features are perfectly
represented by inter-element jumps and high-order basis functions approximate
smooth regions of the solution without nonlinear stabilization, which leads to
accurate approximations on traditionally coarse meshes. In this work, we
introduce a robust implicit shock tracking framework specialized for problems
with parameter-dependent lead shocks (i.e., shocks separating a farfield
condition from the downstream flow), which commonly arise in high-speed
aerodynamics and astrophysics applications. After a shock-aligned mesh is
produced at one parameter configuration, all elements upstream of the lead
shock are removed and the nodes on the lead shock are positioned for new
parameter configurations using the implicit shock tracking solver. The proposed
framework can be used for most many-query applications involving parametrized
lead shocks such as optimization, uncertainty quantification, parameter sweeps,
"what-if" scenarios, or parameter-based continuation. We demonstrate the
robustness and flexibility of the framework using a one-dimensional space-time
Riemann problem, and two- and three-dimensional supersonic and hypersonic
benchmark problems.
| no_new_dataset | 0.711067 |
2304.11431 | Farhad Pourpanah Dr. | Moloud Abdar, Meenakshi Kollati, Swaraja Kuraparthi, Farhad Pourpanah,
Daniel McDuff, Mohammad Ghavamzadeh, Shuicheng Yan, Abduallah Mohamed, Abbas
Khosravi, Erik Cambria, Fatih Porikli | A Review of Deep Learning for Video Captioning | 42 pages, 10 figures | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Video captioning (VC) is a fast-moving, cross-disciplinary area of research
that bridges work in the fields of computer vision, natural language processing
(NLP), linguistics, and human-computer interaction. In essence, VC involves
understanding a video and describing it with language. Captioning is used in a
host of applications from creating more accessible interfaces (e.g., low-vision
navigation) to video question answering (V-QA), video retrieval and content
generation. This survey covers deep learning-based VC, including but, not
limited to, attention-based architectures, graph networks, reinforcement
learning, adversarial networks, dense video captioning (DVC), and more. We
discuss the datasets and evaluation metrics used in the field, and limitations,
applications, challenges, and future directions for VC.
| [
{
"version": "v1",
"created": "Sat, 22 Apr 2023 15:30:54 GMT"
}
] | 2023-04-25T00:00:00 | [
[
"Abdar",
"Moloud",
""
],
[
"Kollati",
"Meenakshi",
""
],
[
"Kuraparthi",
"Swaraja",
""
],
[
"Pourpanah",
"Farhad",
""
],
[
"McDuff",
"Daniel",
""
],
[
"Ghavamzadeh",
"Mohammad",
""
],
[
"Yan",
"Shuicheng",
""
],
[
"Mohamed",
"Abduallah",
""
],
[
"Khosravi",
"Abbas",
""
],
[
"Cambria",
"Erik",
""
],
[
"Porikli",
"Fatih",
""
]
] | TITLE: A Review of Deep Learning for Video Captioning
ABSTRACT: Video captioning (VC) is a fast-moving, cross-disciplinary area of research
that bridges work in the fields of computer vision, natural language processing
(NLP), linguistics, and human-computer interaction. In essence, VC involves
understanding a video and describing it with language. Captioning is used in a
host of applications from creating more accessible interfaces (e.g., low-vision
navigation) to video question answering (V-QA), video retrieval and content
generation. This survey covers deep learning-based VC, including but, not
limited to, attention-based architectures, graph networks, reinforcement
learning, adversarial networks, dense video captioning (DVC), and more. We
discuss the datasets and evaluation metrics used in the field, and limitations,
applications, challenges, and future directions for VC.
| no_new_dataset | 0.709648 |
2304.11432 | Peng Chen | Peng Chen, Xin Du, Zhihui Lu and Hongfeng Chai | Universal Adversarial Backdoor Attacks to Fool Vertical Federated
Learning in Cloud-Edge Collaboration | 14 pages, 7 figures | null | null | null | cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Vertical federated learning (VFL) is a cloud-edge collaboration paradigm that
enables edge nodes, comprising resource-constrained Internet of Things (IoT)
devices, to cooperatively train artificial intelligence (AI) models while
retaining their data locally. This paradigm facilitates improved privacy and
security for edges and IoT devices, making VFL an essential component of
Artificial Intelligence of Things (AIoT) systems. Nevertheless, the partitioned
structure of VFL can be exploited by adversaries to inject a backdoor, enabling
them to manipulate the VFL predictions. In this paper, we aim to investigate
the vulnerability of VFL in the context of binary classification tasks. To this
end, we define a threat model for backdoor attacks in VFL and introduce a
universal adversarial backdoor (UAB) attack to poison the predictions of VFL.
The UAB attack, consisting of universal trigger generation and clean-label
backdoor injection, is incorporated during the VFL training at specific
iterations. This is achieved by alternately optimizing the universal trigger
and model parameters of VFL sub-problems. Our work distinguishes itself from
existing studies on designing backdoor attacks for VFL, as those require the
knowledge of auxiliary information not accessible within the split VFL
architecture. In contrast, our approach does not necessitate any additional
data to execute the attack. On the LendingClub and Zhongyuan datasets, our
approach surpasses existing state-of-the-art methods, achieving up to 100\%
backdoor task performance while maintaining the main task performance. Our
results in this paper make a major advance to revealing the hidden backdoor
risks of VFL, hence paving the way for the future development of secure AIoT.
| [
{
"version": "v1",
"created": "Sat, 22 Apr 2023 15:31:15 GMT"
}
] | 2023-04-25T00:00:00 | [
[
"Chen",
"Peng",
""
],
[
"Du",
"Xin",
""
],
[
"Lu",
"Zhihui",
""
],
[
"Chai",
"Hongfeng",
""
]
] | TITLE: Universal Adversarial Backdoor Attacks to Fool Vertical Federated
Learning in Cloud-Edge Collaboration
ABSTRACT: Vertical federated learning (VFL) is a cloud-edge collaboration paradigm that
enables edge nodes, comprising resource-constrained Internet of Things (IoT)
devices, to cooperatively train artificial intelligence (AI) models while
retaining their data locally. This paradigm facilitates improved privacy and
security for edges and IoT devices, making VFL an essential component of
Artificial Intelligence of Things (AIoT) systems. Nevertheless, the partitioned
structure of VFL can be exploited by adversaries to inject a backdoor, enabling
them to manipulate the VFL predictions. In this paper, we aim to investigate
the vulnerability of VFL in the context of binary classification tasks. To this
end, we define a threat model for backdoor attacks in VFL and introduce a
universal adversarial backdoor (UAB) attack to poison the predictions of VFL.
The UAB attack, consisting of universal trigger generation and clean-label
backdoor injection, is incorporated during the VFL training at specific
iterations. This is achieved by alternately optimizing the universal trigger
and model parameters of VFL sub-problems. Our work distinguishes itself from
existing studies on designing backdoor attacks for VFL, as those require the
knowledge of auxiliary information not accessible within the split VFL
architecture. In contrast, our approach does not necessitate any additional
data to execute the attack. On the LendingClub and Zhongyuan datasets, our
approach surpasses existing state-of-the-art methods, achieving up to 100\%
backdoor task performance while maintaining the main task performance. Our
results in this paper make a major advance to revealing the hidden backdoor
risks of VFL, hence paving the way for the future development of secure AIoT.
| no_new_dataset | 0.708855 |
2304.11433 | Yu Wang | Yu Wang, Zhiwei Liu, Liangwei Yang, Philip S. Yu | Conditional Denoising Diffusion for Sequential Recommendation | null | null | null | null | cs.LG cs.AI cs.IR | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Generative models have attracted significant interest due to their ability to
handle uncertainty by learning the inherent data distributions. However, two
prominent generative models, namely Generative Adversarial Networks (GANs) and
Variational AutoEncoders (VAEs), exhibit challenges that impede achieving
optimal performance in sequential recommendation tasks. Specifically, GANs
suffer from unstable optimization, while VAEs are prone to posterior collapse
and over-smoothed generations. The sparse and noisy nature of sequential
recommendation further exacerbates these issues. In response to these
limitations, we present a conditional denoising diffusion model, which includes
a sequence encoder, a cross-attentive denoising decoder, and a step-wise
diffuser. This approach streamlines the optimization and generation process by
dividing it into easier and tractable steps in a conditional autoregressive
manner. Furthermore, we introduce a novel optimization schema that incorporates
both cross-divergence loss and contrastive loss. This novel training schema
enables the model to generate high-quality sequence/item representations and
meanwhile precluding collapse. We conducted comprehensive experiments on four
benchmark datasets, and the superior performance achieved by our model attests
to its efficacy.
| [
{
"version": "v1",
"created": "Sat, 22 Apr 2023 15:32:59 GMT"
}
] | 2023-04-25T00:00:00 | [
[
"Wang",
"Yu",
""
],
[
"Liu",
"Zhiwei",
""
],
[
"Yang",
"Liangwei",
""
],
[
"Yu",
"Philip S.",
""
]
] | TITLE: Conditional Denoising Diffusion for Sequential Recommendation
ABSTRACT: Generative models have attracted significant interest due to their ability to
handle uncertainty by learning the inherent data distributions. However, two
prominent generative models, namely Generative Adversarial Networks (GANs) and
Variational AutoEncoders (VAEs), exhibit challenges that impede achieving
optimal performance in sequential recommendation tasks. Specifically, GANs
suffer from unstable optimization, while VAEs are prone to posterior collapse
and over-smoothed generations. The sparse and noisy nature of sequential
recommendation further exacerbates these issues. In response to these
limitations, we present a conditional denoising diffusion model, which includes
a sequence encoder, a cross-attentive denoising decoder, and a step-wise
diffuser. This approach streamlines the optimization and generation process by
dividing it into easier and tractable steps in a conditional autoregressive
manner. Furthermore, we introduce a novel optimization schema that incorporates
both cross-divergence loss and contrastive loss. This novel training schema
enables the model to generate high-quality sequence/item representations and
meanwhile precluding collapse. We conducted comprehensive experiments on four
benchmark datasets, and the superior performance achieved by our model attests
to its efficacy.
| no_new_dataset | 0.708975 |
2304.11434 | Raviraj Joshi | Samruddhi Deode, Janhavi Gadre, Aditi Kajale, Ananya Joshi, Raviraj
Joshi | L3Cube-IndicSBERT: A simple approach for learning cross-lingual sentence
representations using multilingual BERT | null | null | null | null | cs.CL cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The multilingual Sentence-BERT (SBERT) models map different languages to
common representation space and are useful for cross-language similarity and
mining tasks. We propose a simple yet effective approach to convert vanilla
multilingual BERT models into multilingual sentence BERT models using synthetic
corpus. We simply aggregate translated NLI or STS datasets of the low-resource
target languages together and perform SBERT-like fine-tuning of the vanilla
multilingual BERT model. We show that multilingual BERT models are inherent
cross-lingual learners and this simple baseline fine-tuning approach without
explicit cross-lingual training yields exceptional cross-lingual properties. We
show the efficacy of our approach on 10 major Indic languages and also show the
applicability of our approach to non-Indic languages German and French. Using
this approach, we further present L3Cube-IndicSBERT, the first multilingual
sentence representation model specifically for Indian languages Hindi, Marathi,
Kannada, Telugu, Malayalam, Tamil, Gujarati, Odia, Bengali, and Punjabi. The
IndicSBERT exhibits strong cross-lingual capabilities and performs
significantly better than alternatives like LaBSE, LASER, and
paraphrase-multilingual-mpnet-base-v2 on Indic cross-lingual and monolingual
sentence similarity tasks. We also release monolingual SBERT models for each of
the languages and show that IndicSBERT performs competitively with its
monolingual counterparts. These models have been evaluated using embedding
similarity scores and classification accuracy.
| [
{
"version": "v1",
"created": "Sat, 22 Apr 2023 15:45:40 GMT"
}
] | 2023-04-25T00:00:00 | [
[
"Deode",
"Samruddhi",
""
],
[
"Gadre",
"Janhavi",
""
],
[
"Kajale",
"Aditi",
""
],
[
"Joshi",
"Ananya",
""
],
[
"Joshi",
"Raviraj",
""
]
] | TITLE: L3Cube-IndicSBERT: A simple approach for learning cross-lingual sentence
representations using multilingual BERT
ABSTRACT: The multilingual Sentence-BERT (SBERT) models map different languages to
common representation space and are useful for cross-language similarity and
mining tasks. We propose a simple yet effective approach to convert vanilla
multilingual BERT models into multilingual sentence BERT models using synthetic
corpus. We simply aggregate translated NLI or STS datasets of the low-resource
target languages together and perform SBERT-like fine-tuning of the vanilla
multilingual BERT model. We show that multilingual BERT models are inherent
cross-lingual learners and this simple baseline fine-tuning approach without
explicit cross-lingual training yields exceptional cross-lingual properties. We
show the efficacy of our approach on 10 major Indic languages and also show the
applicability of our approach to non-Indic languages German and French. Using
this approach, we further present L3Cube-IndicSBERT, the first multilingual
sentence representation model specifically for Indian languages Hindi, Marathi,
Kannada, Telugu, Malayalam, Tamil, Gujarati, Odia, Bengali, and Punjabi. The
IndicSBERT exhibits strong cross-lingual capabilities and performs
significantly better than alternatives like LaBSE, LASER, and
paraphrase-multilingual-mpnet-base-v2 on Indic cross-lingual and monolingual
sentence similarity tasks. We also release monolingual SBERT models for each of
the languages and show that IndicSBERT performs competitively with its
monolingual counterparts. These models have been evaluated using embedding
similarity scores and classification accuracy.
| no_new_dataset | 0.708055 |
2304.11445 | Kudaibergen Abutalip | Kudaibergen Abutalip, Numan Saeed, Mustaqeem Khan, Abdulmotaleb El
Saddik | Improving Stain Invariance of CNNs for Segmentation by Fusing Channel
Attention and Domain-Adversarial Training | null | null | null | null | eess.IV cs.CV | http://creativecommons.org/licenses/by/4.0/ | Variability in staining protocols, such as different slide preparation
techniques, chemicals, and scanner configurations, can result in a diverse set
of whole slide images (WSIs). This distribution shift can negatively impact the
performance of deep learning models on unseen samples, presenting a significant
challenge for developing new computational pathology applications. In this
study, we propose a method for improving the generalizability of convolutional
neural networks (CNNs) to stain changes in a single-source setting for semantic
segmentation. Recent studies indicate that style features mainly exist as
covariances in earlier network layers. We design a channel attention mechanism
based on these findings that detects stain-specific features and modify the
previously proposed stain-invariant training scheme. We reweigh the outputs of
earlier layers and pass them to the stain-adversarial training branch. We
evaluate our method on multi-center, multi-stain datasets and demonstrate its
effectiveness through interpretability analysis. Our approach achieves
substantial improvements over baselines and competitive performance compared to
other methods, as measured by various evaluation metrics. We also show that
combining our method with stain augmentation leads to mutually beneficial
results and outperforms other techniques. Overall, our study makes significant
contributions to the field of computational pathology.
| [
{
"version": "v1",
"created": "Sat, 22 Apr 2023 16:54:37 GMT"
}
] | 2023-04-25T00:00:00 | [
[
"Abutalip",
"Kudaibergen",
""
],
[
"Saeed",
"Numan",
""
],
[
"Khan",
"Mustaqeem",
""
],
[
"Saddik",
"Abdulmotaleb El",
""
]
] | TITLE: Improving Stain Invariance of CNNs for Segmentation by Fusing Channel
Attention and Domain-Adversarial Training
ABSTRACT: Variability in staining protocols, such as different slide preparation
techniques, chemicals, and scanner configurations, can result in a diverse set
of whole slide images (WSIs). This distribution shift can negatively impact the
performance of deep learning models on unseen samples, presenting a significant
challenge for developing new computational pathology applications. In this
study, we propose a method for improving the generalizability of convolutional
neural networks (CNNs) to stain changes in a single-source setting for semantic
segmentation. Recent studies indicate that style features mainly exist as
covariances in earlier network layers. We design a channel attention mechanism
based on these findings that detects stain-specific features and modify the
previously proposed stain-invariant training scheme. We reweigh the outputs of
earlier layers and pass them to the stain-adversarial training branch. We
evaluate our method on multi-center, multi-stain datasets and demonstrate its
effectiveness through interpretability analysis. Our approach achieves
substantial improvements over baselines and competitive performance compared to
other methods, as measured by various evaluation metrics. We also show that
combining our method with stain augmentation leads to mutually beneficial
results and outperforms other techniques. Overall, our study makes significant
contributions to the field of computational pathology.
| no_new_dataset | 0.709239 |
2304.11446 | Yansong Gao Mr. | Yansong Gao, Zhihong Pan, Xin Zhou, Le Kang, Pratik Chaudhari | Fast Diffusion Probabilistic Model Sampling through the lens of Backward
Error Analysis | arXiv admin note: text overlap with arXiv:2101.12176 by other authors | null | null | null | cs.CV cs.AI | http://creativecommons.org/licenses/by/4.0/ | Denoising diffusion probabilistic models (DDPMs) are a class of powerful
generative models. The past few years have witnessed the great success of DDPMs
in generating high-fidelity samples. A significant limitation of the DDPMs is
the slow sampling procedure. DDPMs generally need hundreds or thousands of
sequential function evaluations (steps) of neural networks to generate a
sample. This paper aims to develop a fast sampling method for DDPMs requiring
much fewer steps while retaining high sample quality. The inference process of
DDPMs approximates solving the corresponding diffusion ordinary differential
equations (diffusion ODEs) in the continuous limit. This work analyzes how the
backward error affects the diffusion ODEs and the sample quality in DDPMs. We
propose fast sampling through the \textbf{Restricting Backward Error schedule
(RBE schedule)} based on dynamically moderating the long-time backward error.
Our method accelerates DDPMs without any further training. Our experiments show
that sampling with an RBE schedule generates high-quality samples within only 8
to 20 function evaluations on various benchmark datasets. We achieved 12.01 FID
in 8 function evaluations on the ImageNet $128\times128$, and a $20\times$
speedup compared with previous baseline samplers.
| [
{
"version": "v1",
"created": "Sat, 22 Apr 2023 16:58:47 GMT"
}
] | 2023-04-25T00:00:00 | [
[
"Gao",
"Yansong",
""
],
[
"Pan",
"Zhihong",
""
],
[
"Zhou",
"Xin",
""
],
[
"Kang",
"Le",
""
],
[
"Chaudhari",
"Pratik",
""
]
] | TITLE: Fast Diffusion Probabilistic Model Sampling through the lens of Backward
Error Analysis
ABSTRACT: Denoising diffusion probabilistic models (DDPMs) are a class of powerful
generative models. The past few years have witnessed the great success of DDPMs
in generating high-fidelity samples. A significant limitation of the DDPMs is
the slow sampling procedure. DDPMs generally need hundreds or thousands of
sequential function evaluations (steps) of neural networks to generate a
sample. This paper aims to develop a fast sampling method for DDPMs requiring
much fewer steps while retaining high sample quality. The inference process of
DDPMs approximates solving the corresponding diffusion ordinary differential
equations (diffusion ODEs) in the continuous limit. This work analyzes how the
backward error affects the diffusion ODEs and the sample quality in DDPMs. We
propose fast sampling through the \textbf{Restricting Backward Error schedule
(RBE schedule)} based on dynamically moderating the long-time backward error.
Our method accelerates DDPMs without any further training. Our experiments show
that sampling with an RBE schedule generates high-quality samples within only 8
to 20 function evaluations on various benchmark datasets. We achieved 12.01 FID
in 8 function evaluations on the ImageNet $128\times128$, and a $20\times$
speedup compared with previous baseline samplers.
| no_new_dataset | 0.710848 |
2304.11450 | Omid Nejati Manzari | Davoud Saadati, Omid Nejati Manzari, Sattar Mirzakuchaki | Dilated-UNet: A Fast and Accurate Medical Image Segmentation Approach
using a Dilated Transformer and U-Net Architecture | null | null | null | null | cs.CV | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Medical image segmentation is crucial for the development of computer-aided
diagnostic and therapeutic systems, but still faces numerous difficulties. In
recent years, the commonly used encoder-decoder architecture based on CNNs has
been applied effectively in medical image segmentation, but has limitations in
terms of learning global context and spatial relationships. Some researchers
have attempted to incorporate transformers into both the decoder and encoder
components, with promising results, but this approach still requires further
improvement due to its high computational complexity. This paper introduces
Dilated-UNet, which combines a Dilated Transformer block with the U-Net
architecture for accurate and fast medical image segmentation. Image patches
are transformed into tokens and fed into the U-shaped encoder-decoder
architecture, with skip-connections for local-global semantic feature learning.
The encoder uses a hierarchical Dilated Transformer with a combination of
Neighborhood Attention and Dilated Neighborhood Attention Transformer to
extract local and sparse global attention. The results of our experiments show
that Dilated-UNet outperforms other models on several challenging medical image
segmentation datasets, such as ISIC and Synapse.
| [
{
"version": "v1",
"created": "Sat, 22 Apr 2023 17:20:13 GMT"
}
] | 2023-04-25T00:00:00 | [
[
"Saadati",
"Davoud",
""
],
[
"Manzari",
"Omid Nejati",
""
],
[
"Mirzakuchaki",
"Sattar",
""
]
] | TITLE: Dilated-UNet: A Fast and Accurate Medical Image Segmentation Approach
using a Dilated Transformer and U-Net Architecture
ABSTRACT: Medical image segmentation is crucial for the development of computer-aided
diagnostic and therapeutic systems, but still faces numerous difficulties. In
recent years, the commonly used encoder-decoder architecture based on CNNs has
been applied effectively in medical image segmentation, but has limitations in
terms of learning global context and spatial relationships. Some researchers
have attempted to incorporate transformers into both the decoder and encoder
components, with promising results, but this approach still requires further
improvement due to its high computational complexity. This paper introduces
Dilated-UNet, which combines a Dilated Transformer block with the U-Net
architecture for accurate and fast medical image segmentation. Image patches
are transformed into tokens and fed into the U-shaped encoder-decoder
architecture, with skip-connections for local-global semantic feature learning.
The encoder uses a hierarchical Dilated Transformer with a combination of
Neighborhood Attention and Dilated Neighborhood Attention Transformer to
extract local and sparse global attention. The results of our experiments show
that Dilated-UNet outperforms other models on several challenging medical image
segmentation datasets, such as ISIC and Synapse.
| no_new_dataset | 0.709806 |
2304.11470 | Haotian Xue | Haotian Xue, Antonio Torralba, Joshua B. Tenenbaum, Daniel LK Yamins,
Yunzhu Li, Hsiao-Yu Tung | 3D-IntPhys: Towards More Generalized 3D-grounded Visual Intuitive
Physics under Challenging Scenes | null | null | null | null | cs.CV cs.AI | http://creativecommons.org/licenses/by/4.0/ | Given a visual scene, humans have strong intuitions about how a scene can
evolve over time under given actions. The intuition, often termed visual
intuitive physics, is a critical ability that allows us to make effective plans
to manipulate the scene to achieve desired outcomes without relying on
extensive trial and error. In this paper, we present a framework capable of
learning 3D-grounded visual intuitive physics models from videos of complex
scenes with fluids. Our method is composed of a conditional Neural Radiance
Field (NeRF)-style visual frontend and a 3D point-based dynamics prediction
backend, using which we can impose strong relational and structural inductive
bias to capture the structure of the underlying environment. Unlike existing
intuitive point-based dynamics works that rely on the supervision of dense
point trajectory from simulators, we relax the requirements and only assume
access to multi-view RGB images and (imperfect) instance masks acquired using
color prior. This enables the proposed model to handle scenarios where accurate
point estimation and tracking are hard or impossible. We generate datasets
including three challenging scenarios involving fluid, granular materials, and
rigid objects in the simulation. The datasets do not include any dense particle
information so most previous 3D-based intuitive physics pipelines can barely
deal with that. We show our model can make long-horizon future predictions by
learning from raw images and significantly outperforms models that do not
employ an explicit 3D representation space. We also show that once trained, our
model can achieve strong generalization in complex scenarios under extrapolate
settings.
| [
{
"version": "v1",
"created": "Sat, 22 Apr 2023 19:28:49 GMT"
}
] | 2023-04-25T00:00:00 | [
[
"Xue",
"Haotian",
""
],
[
"Torralba",
"Antonio",
""
],
[
"Tenenbaum",
"Joshua B.",
""
],
[
"Yamins",
"Daniel LK",
""
],
[
"Li",
"Yunzhu",
""
],
[
"Tung",
"Hsiao-Yu",
""
]
] | TITLE: 3D-IntPhys: Towards More Generalized 3D-grounded Visual Intuitive
Physics under Challenging Scenes
ABSTRACT: Given a visual scene, humans have strong intuitions about how a scene can
evolve over time under given actions. The intuition, often termed visual
intuitive physics, is a critical ability that allows us to make effective plans
to manipulate the scene to achieve desired outcomes without relying on
extensive trial and error. In this paper, we present a framework capable of
learning 3D-grounded visual intuitive physics models from videos of complex
scenes with fluids. Our method is composed of a conditional Neural Radiance
Field (NeRF)-style visual frontend and a 3D point-based dynamics prediction
backend, using which we can impose strong relational and structural inductive
bias to capture the structure of the underlying environment. Unlike existing
intuitive point-based dynamics works that rely on the supervision of dense
point trajectory from simulators, we relax the requirements and only assume
access to multi-view RGB images and (imperfect) instance masks acquired using
color prior. This enables the proposed model to handle scenarios where accurate
point estimation and tracking are hard or impossible. We generate datasets
including three challenging scenarios involving fluid, granular materials, and
rigid objects in the simulation. The datasets do not include any dense particle
information so most previous 3D-based intuitive physics pipelines can barely
deal with that. We show our model can make long-horizon future predictions by
learning from raw images and significantly outperforms models that do not
employ an explicit 3D representation space. We also show that once trained, our
model can achieve strong generalization in complex scenarios under extrapolate
settings.
| no_new_dataset | 0.710635 |
2304.11472 | Orchid Chetia Phukan | Orchid Chetia Phukan, Arun Balaji Buduru, Rajesh Sharma | A Comparative Study of Pre-trained Speech and Audio Embeddings for
Speech Emotion Recognition | null | null | null | null | eess.AS cs.AI cs.LG | http://creativecommons.org/licenses/by/4.0/ | Pre-trained models (PTMs) have shown great promise in the speech and audio
domain. Embeddings leveraged from these models serve as inputs for learning
algorithms with applications in various downstream tasks. One such crucial task
is Speech Emotion Recognition (SER) which has a wide range of applications,
including dynamic analysis of customer calls, mental health assessment, and
personalized language learning. PTM embeddings have helped advance SER,
however, a comprehensive comparison of these PTM embeddings that consider
multiple facets such as embedding model architecture, data used for
pre-training, and the pre-training procedure being followed is missing. A
thorough comparison of PTM embeddings will aid in the faster and more efficient
development of models and enable their deployment in real-world scenarios. In
this work, we exploit this research gap and perform a comparative analysis of
embeddings extracted from eight speech and audio PTMs (wav2vec 2.0, data2vec,
wavLM, UniSpeech-SAT, wav2clip, YAMNet, x-vector, ECAPA). We perform an
extensive empirical analysis with four speech emotion datasets (CREMA-D, TESS,
SAVEE, Emo-DB) by training three algorithms (XGBoost, Random Forest, FCN) on
the derived embeddings. The results of our study indicate that the best
performance is achieved by algorithms trained on embeddings derived from PTMs
trained for speaker recognition followed by wav2clip and UniSpeech-SAT. This
can relay that the top performance by embeddings from speaker recognition PTMs
is most likely due to the model taking up information about numerous speech
features such as tone, accent, pitch, and so on during its speaker recognition
training. Insights from this work will assist future studies in their selection
of embeddings for applications related to SER.
| [
{
"version": "v1",
"created": "Sat, 22 Apr 2023 19:56:35 GMT"
}
] | 2023-04-25T00:00:00 | [
[
"Phukan",
"Orchid Chetia",
""
],
[
"Buduru",
"Arun Balaji",
""
],
[
"Sharma",
"Rajesh",
""
]
] | TITLE: A Comparative Study of Pre-trained Speech and Audio Embeddings for
Speech Emotion Recognition
ABSTRACT: Pre-trained models (PTMs) have shown great promise in the speech and audio
domain. Embeddings leveraged from these models serve as inputs for learning
algorithms with applications in various downstream tasks. One such crucial task
is Speech Emotion Recognition (SER) which has a wide range of applications,
including dynamic analysis of customer calls, mental health assessment, and
personalized language learning. PTM embeddings have helped advance SER,
however, a comprehensive comparison of these PTM embeddings that consider
multiple facets such as embedding model architecture, data used for
pre-training, and the pre-training procedure being followed is missing. A
thorough comparison of PTM embeddings will aid in the faster and more efficient
development of models and enable their deployment in real-world scenarios. In
this work, we exploit this research gap and perform a comparative analysis of
embeddings extracted from eight speech and audio PTMs (wav2vec 2.0, data2vec,
wavLM, UniSpeech-SAT, wav2clip, YAMNet, x-vector, ECAPA). We perform an
extensive empirical analysis with four speech emotion datasets (CREMA-D, TESS,
SAVEE, Emo-DB) by training three algorithms (XGBoost, Random Forest, FCN) on
the derived embeddings. The results of our study indicate that the best
performance is achieved by algorithms trained on embeddings derived from PTMs
trained for speaker recognition followed by wav2clip and UniSpeech-SAT. This
can relay that the top performance by embeddings from speaker recognition PTMs
is most likely due to the model taking up information about numerous speech
features such as tone, accent, pitch, and so on during its speaker recognition
training. Insights from this work will assist future studies in their selection
of embeddings for applications related to SER.
| no_new_dataset | 0.709057 |
2304.11479 | Eunseop Lee | Eunseop Lee, Inhan Kim and Daijin Kim | Weight-based Mask for Domain Adaptation | 5 pages, 2 figures, to be published in ICASSP2023 | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In computer vision, unsupervised domain adaptation (UDA) is an approach to
transferring knowledge from a label-rich source domain to a fully-unlabeled
target domain. Conventional UDA approaches have two problems. The first problem
is that a class classifier can be biased to the source domain because it is
trained using only source samples. The second is that previous approaches align
image-level features regardless of foreground and background, although the
classifier requires foreground features. To solve these problems, we introduce
Weight-based Mask Network (WEMNet) composed of Domain Ignore Module (DIM) and
Semantic Enhancement Module (SEM). DIM obtains domain-agnostic feature
representations via the weight of the domain discriminator and predicts
categories. In addition, SEM obtains class-related feature representations
using the classifier weight and focuses on the foreground features for domain
adaptation. Extensive experimental results reveal that the proposed WEMNet
outperforms the competitive accuracy on representative UDA datasets.
| [
{
"version": "v1",
"created": "Sat, 22 Apr 2023 20:51:07 GMT"
}
] | 2023-04-25T00:00:00 | [
[
"Lee",
"Eunseop",
""
],
[
"Kim",
"Inhan",
""
],
[
"Kim",
"Daijin",
""
]
] | TITLE: Weight-based Mask for Domain Adaptation
ABSTRACT: In computer vision, unsupervised domain adaptation (UDA) is an approach to
transferring knowledge from a label-rich source domain to a fully-unlabeled
target domain. Conventional UDA approaches have two problems. The first problem
is that a class classifier can be biased to the source domain because it is
trained using only source samples. The second is that previous approaches align
image-level features regardless of foreground and background, although the
classifier requires foreground features. To solve these problems, we introduce
Weight-based Mask Network (WEMNet) composed of Domain Ignore Module (DIM) and
Semantic Enhancement Module (SEM). DIM obtains domain-agnostic feature
representations via the weight of the domain discriminator and predicts
categories. In addition, SEM obtains class-related feature representations
using the classifier weight and focuses on the foreground features for domain
adaptation. Extensive experimental results reveal that the proposed WEMNet
outperforms the competitive accuracy on representative UDA datasets.
| no_new_dataset | 0.710226 |
2304.11513 | Yue Hu | Yue Hu, Yuhang Zhang, Yanbing Wang, Daniel Work | Detecting Socially Abnormal Highway Driving Behaviors via Recurrent
Graph Attention Networks | null | null | 10.1145/3543507.3583452. | null | cs.AI | http://creativecommons.org/licenses/by/4.0/ | With the rapid development of Internet of Things technologies, the next
generation traffic monitoring infrastructures are connected via the web, to aid
traffic data collection and intelligent traffic management. One of the most
important tasks in traffic is anomaly detection, since abnormal drivers can
reduce traffic efficiency and cause safety issues. This work focuses on
detecting abnormal driving behaviors from trajectories produced by highway
video surveillance systems. Most of the current abnormal driving behavior
detection methods focus on a limited category of abnormal behaviors that deal
with a single vehicle without considering vehicular interactions. In this work,
we consider the problem of detecting a variety of socially abnormal driving
behaviors, i.e., behaviors that do not conform to the behavior of other nearby
drivers. This task is complicated by the variety of vehicular interactions and
the spatial-temporal varying nature of highway traffic. To solve this problem,
we propose an autoencoder with a Recurrent Graph Attention Network that can
capture the highway driving behaviors contextualized on the surrounding cars,
and detect anomalies that deviate from learned patterns. Our model is scalable
to large freeways with thousands of cars. Experiments on data generated from
traffic simulation software show that our model is the only one that can spot
the exact vehicle conducting socially abnormal behaviors, among the
state-of-the-art anomaly detection models. We further show the performance on
real world HighD traffic dataset, where our model detects vehicles that violate
the local driving norms.
| [
{
"version": "v1",
"created": "Sun, 23 Apr 2023 01:32:47 GMT"
}
] | 2023-04-25T00:00:00 | [
[
"Hu",
"Yue",
""
],
[
"Zhang",
"Yuhang",
""
],
[
"Wang",
"Yanbing",
""
],
[
"Work",
"Daniel",
""
]
] | TITLE: Detecting Socially Abnormal Highway Driving Behaviors via Recurrent
Graph Attention Networks
ABSTRACT: With the rapid development of Internet of Things technologies, the next
generation traffic monitoring infrastructures are connected via the web, to aid
traffic data collection and intelligent traffic management. One of the most
important tasks in traffic is anomaly detection, since abnormal drivers can
reduce traffic efficiency and cause safety issues. This work focuses on
detecting abnormal driving behaviors from trajectories produced by highway
video surveillance systems. Most of the current abnormal driving behavior
detection methods focus on a limited category of abnormal behaviors that deal
with a single vehicle without considering vehicular interactions. In this work,
we consider the problem of detecting a variety of socially abnormal driving
behaviors, i.e., behaviors that do not conform to the behavior of other nearby
drivers. This task is complicated by the variety of vehicular interactions and
the spatial-temporal varying nature of highway traffic. To solve this problem,
we propose an autoencoder with a Recurrent Graph Attention Network that can
capture the highway driving behaviors contextualized on the surrounding cars,
and detect anomalies that deviate from learned patterns. Our model is scalable
to large freeways with thousands of cars. Experiments on data generated from
traffic simulation software show that our model is the only one that can spot
the exact vehicle conducting socially abnormal behaviors, among the
state-of-the-art anomaly detection models. We further show the performance on
real world HighD traffic dataset, where our model detects vehicles that violate
the local driving norms.
| no_new_dataset | 0.687997 |
2304.11528 | Jiahao Liu | Jiahao Liu, Dongsheng Li, Hansu Gu, Tun Lu, Peng Zhang, Li Shang and
Ning Gu | Triple Structural Information Modelling for Accurate, Explainable and
Interactive Recommendation | 10 pages, Accepted by SIGIR 2023 | null | null | null | cs.IR cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In dynamic interaction graphs, user-item interactions usually follow
heterogeneous patterns, represented by different structural information, such
as user-item co-occurrence, sequential information of user interactions and the
transition probabilities of item pairs. However, the existing methods cannot
simultaneously leverage all three structural information, resulting in
suboptimal performance. To this end, we propose TriSIM4Rec, a triple structural
information modeling method for accurate, explainable and interactive
recommendation on dynamic interaction graphs. Specifically, TriSIM4Rec consists
of 1) a dynamic ideal low-pass graph filter to dynamically mine co-occurrence
information in user-item interactions, which is implemented by incremental
singular value decomposition (SVD); 2) a parameter-free attention module to
capture sequential information of user interactions effectively and
efficiently; and 3) an item transition matrix to store the transition
probabilities of item pairs. Then, we fuse the predictions from the triple
structural information sources to obtain the final recommendation results. By
analyzing the relationship between the SVD-based and the recently emerging
graph signal processing (GSP)-based collaborative filtering methods, we find
that the essence of SVD is an ideal low-pass graph filter, so that the interest
vector space in TriSIM4Rec can be extended to achieve explainable and
interactive recommendation, making it possible for users to actively break
through the information cocoons. Experiments on six public datasets
demonstrated the effectiveness of TriSIM4Rec in accuracy, explainability and
interactivity.
| [
{
"version": "v1",
"created": "Sun, 23 Apr 2023 04:00:52 GMT"
}
] | 2023-04-25T00:00:00 | [
[
"Liu",
"Jiahao",
""
],
[
"Li",
"Dongsheng",
""
],
[
"Gu",
"Hansu",
""
],
[
"Lu",
"Tun",
""
],
[
"Zhang",
"Peng",
""
],
[
"Shang",
"Li",
""
],
[
"Gu",
"Ning",
""
]
] | TITLE: Triple Structural Information Modelling for Accurate, Explainable and
Interactive Recommendation
ABSTRACT: In dynamic interaction graphs, user-item interactions usually follow
heterogeneous patterns, represented by different structural information, such
as user-item co-occurrence, sequential information of user interactions and the
transition probabilities of item pairs. However, the existing methods cannot
simultaneously leverage all three structural information, resulting in
suboptimal performance. To this end, we propose TriSIM4Rec, a triple structural
information modeling method for accurate, explainable and interactive
recommendation on dynamic interaction graphs. Specifically, TriSIM4Rec consists
of 1) a dynamic ideal low-pass graph filter to dynamically mine co-occurrence
information in user-item interactions, which is implemented by incremental
singular value decomposition (SVD); 2) a parameter-free attention module to
capture sequential information of user interactions effectively and
efficiently; and 3) an item transition matrix to store the transition
probabilities of item pairs. Then, we fuse the predictions from the triple
structural information sources to obtain the final recommendation results. By
analyzing the relationship between the SVD-based and the recently emerging
graph signal processing (GSP)-based collaborative filtering methods, we find
that the essence of SVD is an ideal low-pass graph filter, so that the interest
vector space in TriSIM4Rec can be extended to achieve explainable and
interactive recommendation, making it possible for users to actively break
through the information cocoons. Experiments on six public datasets
demonstrated the effectiveness of TriSIM4Rec in accuracy, explainability and
interactivity.
| no_new_dataset | 0.709787 |
2304.11529 | Debesh Jha | Smriti Regmi, Aliza Subedi, Ulas Bagci, Debesh Jha | Vision Transformer for Efficient Chest X-ray and Gastrointestinal Image
Classification | null | null | null | null | eess.IV cs.CV | http://creativecommons.org/licenses/by/4.0/ | Medical image analysis is a hot research topic because of its usefulness in
different clinical applications, such as early disease diagnosis and treatment.
Convolutional neural networks (CNNs) have become the de-facto standard in
medical image analysis tasks because of their ability to learn complex features
from the available datasets, which makes them surpass humans in many
image-understanding tasks. In addition to CNNs, transformer architectures also
have gained popularity for medical image analysis tasks. However, despite
progress in the field, there are still potential areas for improvement. This
study uses different CNNs and transformer-based methods with a wide range of
data augmentation techniques. We evaluated their performance on three medical
image datasets from different modalities. We evaluated and compared the
performance of the vision transformer model with other state-of-the-art (SOTA)
pre-trained CNN networks. For Chest X-ray, our vision transformer model
achieved the highest F1 score of 0.9532, recall of 0.9533, Matthews correlation
coefficient (MCC) of 0.9259, and ROC-AUC score of 0.97. Similarly, for the
Kvasir dataset, we achieved an F1 score of 0.9436, recall of 0.9437, MCC of
0.9360, and ROC-AUC score of 0.97. For the Kvasir-Capsule (a large-scale VCE
dataset), our ViT model achieved a weighted F1-score of 0.7156, recall of
0.7182, MCC of 0.3705, and ROC-AUC score of 0.57. We found that our
transformer-based models were better or more effective than various CNN models
for classifying different anatomical structures, findings, and abnormalities.
Our model showed improvement over the CNN-based approaches and suggests that it
could be used as a new benchmarking algorithm for algorithm development.
| [
{
"version": "v1",
"created": "Sun, 23 Apr 2023 04:07:03 GMT"
}
] | 2023-04-25T00:00:00 | [
[
"Regmi",
"Smriti",
""
],
[
"Subedi",
"Aliza",
""
],
[
"Bagci",
"Ulas",
""
],
[
"Jha",
"Debesh",
""
]
] | TITLE: Vision Transformer for Efficient Chest X-ray and Gastrointestinal Image
Classification
ABSTRACT: Medical image analysis is a hot research topic because of its usefulness in
different clinical applications, such as early disease diagnosis and treatment.
Convolutional neural networks (CNNs) have become the de-facto standard in
medical image analysis tasks because of their ability to learn complex features
from the available datasets, which makes them surpass humans in many
image-understanding tasks. In addition to CNNs, transformer architectures also
have gained popularity for medical image analysis tasks. However, despite
progress in the field, there are still potential areas for improvement. This
study uses different CNNs and transformer-based methods with a wide range of
data augmentation techniques. We evaluated their performance on three medical
image datasets from different modalities. We evaluated and compared the
performance of the vision transformer model with other state-of-the-art (SOTA)
pre-trained CNN networks. For Chest X-ray, our vision transformer model
achieved the highest F1 score of 0.9532, recall of 0.9533, Matthews correlation
coefficient (MCC) of 0.9259, and ROC-AUC score of 0.97. Similarly, for the
Kvasir dataset, we achieved an F1 score of 0.9436, recall of 0.9437, MCC of
0.9360, and ROC-AUC score of 0.97. For the Kvasir-Capsule (a large-scale VCE
dataset), our ViT model achieved a weighted F1-score of 0.7156, recall of
0.7182, MCC of 0.3705, and ROC-AUC score of 0.57. We found that our
transformer-based models were better or more effective than various CNN models
for classifying different anatomical structures, findings, and abnormalities.
Our model showed improvement over the CNN-based approaches and suggests that it
could be used as a new benchmarking algorithm for algorithm development.
| no_new_dataset | 0.708458 |
2304.11539 | Yazhou Yao | Rui Chen, Tao Chen, Qiong Wang, Yazhou Yao | Semi-Supervised Semantic Segmentation With Region Relevance | accepted by IEEE International Conference on Multimedia and Expo 2023 | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Semi-supervised semantic segmentation aims to learn from a small amount of
labeled data and plenty of unlabeled ones for the segmentation task. The most
common approach is to generate pseudo-labels for unlabeled images to augment
the training data. However, the noisy pseudo-labels will lead to cumulative
classification errors and aggravate the local inconsistency in prediction. This
paper proposes a Region Relevance Network (RRN) to alleviate the problem
mentioned above. Specifically, we first introduce a local pseudo-label
filtering module that leverages discriminator networks to assess the accuracy
of the pseudo-label at the region level. A local selection loss is proposed to
mitigate the negative impact of wrong pseudo-labels in consistency
regularization training. In addition, we propose a dynamic region-loss
correction module, which takes the merit of network diversity to further rate
the reliability of pseudo-labels and correct the convergence direction of the
segmentation network with a dynamic region loss. Extensive experiments are
conducted on PASCAL VOC 2012 and Cityscapes datasets with varying amounts of
labeled data, demonstrating that our proposed approach achieves
state-of-the-art performance compared to current counterparts.
| [
{
"version": "v1",
"created": "Sun, 23 Apr 2023 04:51:27 GMT"
}
] | 2023-04-25T00:00:00 | [
[
"Chen",
"Rui",
""
],
[
"Chen",
"Tao",
""
],
[
"Wang",
"Qiong",
""
],
[
"Yao",
"Yazhou",
""
]
] | TITLE: Semi-Supervised Semantic Segmentation With Region Relevance
ABSTRACT: Semi-supervised semantic segmentation aims to learn from a small amount of
labeled data and plenty of unlabeled ones for the segmentation task. The most
common approach is to generate pseudo-labels for unlabeled images to augment
the training data. However, the noisy pseudo-labels will lead to cumulative
classification errors and aggravate the local inconsistency in prediction. This
paper proposes a Region Relevance Network (RRN) to alleviate the problem
mentioned above. Specifically, we first introduce a local pseudo-label
filtering module that leverages discriminator networks to assess the accuracy
of the pseudo-label at the region level. A local selection loss is proposed to
mitigate the negative impact of wrong pseudo-labels in consistency
regularization training. In addition, we propose a dynamic region-loss
correction module, which takes the merit of network diversity to further rate
the reliability of pseudo-labels and correct the convergence direction of the
segmentation network with a dynamic region loss. Extensive experiments are
conducted on PASCAL VOC 2012 and Cityscapes datasets with varying amounts of
labeled data, demonstrating that our proposed approach achieves
state-of-the-art performance compared to current counterparts.
| no_new_dataset | 0.710459 |
2304.11595 | Yongcheng Jing | Yongcheng Jing, Xinchao Wang, Dacheng Tao | Segment Anything in Non-Euclidean Domains: Challenges and Opportunities | Work in progress | null | null | null | cs.CV cs.AI cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The recent work known as Segment Anything (SA) has made significant strides
in pushing the boundaries of semantic segmentation into the era of foundation
models. The impact of SA has sparked extremely active discussions and ushered
in an encouraging new wave of developing foundation models for the diverse
tasks in the Euclidean domain, such as object detection and image inpainting.
Despite the promising advances led by SA, the concept has yet to be extended to
the non-Euclidean graph domain. In this paper, we explore a novel Segment
Non-Euclidean Anything (SNA) paradigm that strives to develop foundation models
that can handle the diverse range of graph data within the non-Euclidean
domain, seeking to expand the scope of SA and lay the groundwork for future
research in this direction. To achieve this goal, we begin by discussing the
recent achievements in foundation models associated with SA. We then shed light
on the unique challenges that arise when applying the SA concept to graph
analysis, which involves understanding the differences between the Euclidean
and non-Euclidean domains from both the data and task perspectives. Motivated
by these observations, we present several preliminary solutions to tackle the
challenges of SNA and detail their corresponding limitations, along with
several potential directions to pave the way for future SNA research.
Experiments on five Open Graph Benchmark (OGB) datasets across various tasks,
including graph property classification and regression, as well as multi-label
prediction, demonstrate that the performance of the naive SNA solutions has
considerable room for improvement, pointing towards a promising avenue for
future exploration of Graph General Intelligence.
| [
{
"version": "v1",
"created": "Sun, 23 Apr 2023 10:01:34 GMT"
}
] | 2023-04-25T00:00:00 | [
[
"Jing",
"Yongcheng",
""
],
[
"Wang",
"Xinchao",
""
],
[
"Tao",
"Dacheng",
""
]
] | TITLE: Segment Anything in Non-Euclidean Domains: Challenges and Opportunities
ABSTRACT: The recent work known as Segment Anything (SA) has made significant strides
in pushing the boundaries of semantic segmentation into the era of foundation
models. The impact of SA has sparked extremely active discussions and ushered
in an encouraging new wave of developing foundation models for the diverse
tasks in the Euclidean domain, such as object detection and image inpainting.
Despite the promising advances led by SA, the concept has yet to be extended to
the non-Euclidean graph domain. In this paper, we explore a novel Segment
Non-Euclidean Anything (SNA) paradigm that strives to develop foundation models
that can handle the diverse range of graph data within the non-Euclidean
domain, seeking to expand the scope of SA and lay the groundwork for future
research in this direction. To achieve this goal, we begin by discussing the
recent achievements in foundation models associated with SA. We then shed light
on the unique challenges that arise when applying the SA concept to graph
analysis, which involves understanding the differences between the Euclidean
and non-Euclidean domains from both the data and task perspectives. Motivated
by these observations, we present several preliminary solutions to tackle the
challenges of SNA and detail their corresponding limitations, along with
several potential directions to pave the way for future SNA research.
Experiments on five Open Graph Benchmark (OGB) datasets across various tasks,
including graph property classification and regression, as well as multi-label
prediction, demonstrate that the performance of the naive SNA solutions has
considerable room for improvement, pointing towards a promising avenue for
future exploration of Graph General Intelligence.
| no_new_dataset | 0.709585 |
2304.11598 | Piotr Koniusz | Hao Zhu and Piotr Koniusz | Transductive Few-shot Learning with Prototype-based Label Propagation by
Iterative Graph Refinement | This paper is published at CVPR 2023 | null | null | null | cs.CV cs.AI cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Few-shot learning (FSL) is popular due to its ability to adapt to novel
classes. Compared with inductive few-shot learning, transductive models
typically perform better as they leverage all samples of the query set. The two
existing classes of methods, prototype-based and graph-based, have the
disadvantages of inaccurate prototype estimation and sub-optimal graph
construction with kernel functions, respectively. In this paper, we propose a
novel prototype-based label propagation to solve these issues. Specifically,
our graph construction is based on the relation between prototypes and samples
rather than between samples. As prototypes are being updated, the graph
changes. We also estimate the label of each prototype instead of considering a
prototype be the class centre. On mini-ImageNet, tiered-ImageNet, CIFAR-FS and
CUB datasets, we show the proposed method outperforms other state-of-the-art
methods in transductive FSL and semi-supervised FSL when some unlabeled data
accompanies the novel few-shot task.
| [
{
"version": "v1",
"created": "Sun, 23 Apr 2023 10:09:26 GMT"
}
] | 2023-04-25T00:00:00 | [
[
"Zhu",
"Hao",
""
],
[
"Koniusz",
"Piotr",
""
]
] | TITLE: Transductive Few-shot Learning with Prototype-based Label Propagation by
Iterative Graph Refinement
ABSTRACT: Few-shot learning (FSL) is popular due to its ability to adapt to novel
classes. Compared with inductive few-shot learning, transductive models
typically perform better as they leverage all samples of the query set. The two
existing classes of methods, prototype-based and graph-based, have the
disadvantages of inaccurate prototype estimation and sub-optimal graph
construction with kernel functions, respectively. In this paper, we propose a
novel prototype-based label propagation to solve these issues. Specifically,
our graph construction is based on the relation between prototypes and samples
rather than between samples. As prototypes are being updated, the graph
changes. We also estimate the label of each prototype instead of considering a
prototype be the class centre. On mini-ImageNet, tiered-ImageNet, CIFAR-FS and
CUB datasets, we show the proposed method outperforms other state-of-the-art
methods in transductive FSL and semi-supervised FSL when some unlabeled data
accompanies the novel few-shot task.
| no_new_dataset | 0.709655 |
2304.11618 | Yichi Zhang | Yichi Zhang, Mingyang Chen, Wen Zhang | Modality-Aware Negative Sampling for Multi-modal Knowledge Graph
Embedding | Accepted by IJCNN2023. Code is released in
https://github.com/zjukg/MANS | null | null | null | cs.CL cs.AI | http://creativecommons.org/licenses/by-sa/4.0/ | Negative sampling (NS) is widely used in knowledge graph embedding (KGE),
which aims to generate negative triples to make a positive-negative contrast
during training. However, existing NS methods are unsuitable when multi-modal
information is considered in KGE models. They are also inefficient due to their
complex design. In this paper, we propose Modality-Aware Negative Sampling
(MANS) for multi-modal knowledge graph embedding (MMKGE) to address the
mentioned problems. MANS could align structural and visual embeddings for
entities in KGs and learn meaningful embeddings to perform better in
multi-modal KGE while keeping lightweight and efficient. Empirical results on
two benchmarks demonstrate that MANS outperforms existing NS methods.
Meanwhile, we make further explorations about MANS to confirm its
effectiveness.
| [
{
"version": "v1",
"created": "Sun, 23 Apr 2023 11:22:17 GMT"
}
] | 2023-04-25T00:00:00 | [
[
"Zhang",
"Yichi",
""
],
[
"Chen",
"Mingyang",
""
],
[
"Zhang",
"Wen",
""
]
] | TITLE: Modality-Aware Negative Sampling for Multi-modal Knowledge Graph
Embedding
ABSTRACT: Negative sampling (NS) is widely used in knowledge graph embedding (KGE),
which aims to generate negative triples to make a positive-negative contrast
during training. However, existing NS methods are unsuitable when multi-modal
information is considered in KGE models. They are also inefficient due to their
complex design. In this paper, we propose Modality-Aware Negative Sampling
(MANS) for multi-modal knowledge graph embedding (MMKGE) to address the
mentioned problems. MANS could align structural and visual embeddings for
entities in KGs and learn meaningful embeddings to perform better in
multi-modal KGE while keeping lightweight and efficient. Empirical results on
two benchmarks demonstrate that MANS outperforms existing NS methods.
Meanwhile, we make further explorations about MANS to confirm its
effectiveness.
| no_new_dataset | 0.709655 |
2304.11619 | Jonathan Roberts | Jonathan Roberts, Kai Han, Samuel Albanie | SATIN: A Multi-Task Metadataset for Classifying Satellite Imagery using
Vision-Language Models | null | null | null | null | cs.CV cs.AI cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Interpreting remote sensing imagery enables numerous downstream applications
ranging from land-use planning to deforestation monitoring. Robustly
classifying this data is challenging due to the Earth's geographic diversity.
While many distinct satellite and aerial image classification datasets exist,
there is yet to be a benchmark curated that suitably covers this diversity. In
this work, we introduce SATellite ImageNet (SATIN), a metadataset curated from
27 existing remotely sensed datasets, and comprehensively evaluate the
zero-shot transfer classification capabilities of a broad range of
vision-language (VL) models on SATIN. We find SATIN to be a challenging
benchmark-the strongest method we evaluate achieves a classification accuracy
of 52.0%. We provide a $\href{https://satinbenchmark.github.io}{\text{public
leaderboard}}$ to guide and track the progress of VL models in this important
domain.
| [
{
"version": "v1",
"created": "Sun, 23 Apr 2023 11:23:05 GMT"
}
] | 2023-04-25T00:00:00 | [
[
"Roberts",
"Jonathan",
""
],
[
"Han",
"Kai",
""
],
[
"Albanie",
"Samuel",
""
]
] | TITLE: SATIN: A Multi-Task Metadataset for Classifying Satellite Imagery using
Vision-Language Models
ABSTRACT: Interpreting remote sensing imagery enables numerous downstream applications
ranging from land-use planning to deforestation monitoring. Robustly
classifying this data is challenging due to the Earth's geographic diversity.
While many distinct satellite and aerial image classification datasets exist,
there is yet to be a benchmark curated that suitably covers this diversity. In
this work, we introduce SATellite ImageNet (SATIN), a metadataset curated from
27 existing remotely sensed datasets, and comprehensively evaluate the
zero-shot transfer classification capabilities of a broad range of
vision-language (VL) models on SATIN. We find SATIN to be a challenging
benchmark-the strongest method we evaluate achieves a classification accuracy
of 52.0%. We provide a $\href{https://satinbenchmark.github.io}{\text{public
leaderboard}}$ to guide and track the progress of VL models in this important
domain.
| new_dataset | 0.71584 |
2304.11628 | Zhibing Li Bin | Tinghui Chen, Shuai Li | Using Alternation Direction Method of Multipliers to Enhance robots
Calibration Accuracy based on Multi-Planal Constraints | 7 | null | null | null | cs.RO eess.SP | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | With the widespread application of industrial robots, the problem of absolute
positioning accuracy becomes increasingly prominent. To ensure the working
state of the robots, researchers commonly adopt calibration techniques to
improve its accuracy. However, an industrial robot's working space is mostly
restricted in real working environments, making the collected samples fail in
covering the actual working space to result in the overall migration data. To
address this vital issue, this work proposes a novel industrial robot
calibrator that integrates a measurement configurations selection (MCS) method
and an alternation-direction-method-of-multipliers with multiple planes
constraints (AMPC) algorithm into its working process, whose ideas are
three-fold: a) selecting a group of optimal measurement configurations based on
the observability index to suppress the measurement noises, b) developing an
AMPC algorithm that evidently enhances the calibration accuracy and suppresses
the long-tail convergence, and c) proposing an industrial robot calibration
algorithm that incorporates MCS and AMPC to optimize an industrial robot's
kinematic parameters efficiently. For validating its performance, a
public-available dataset (HRS-P) is established on an HRS-JR680 industrial
robot. Extensive experimental results demonstrate that the proposed calibrator
outperforms several state-of-the-art models in calibration accuracy.
| [
{
"version": "v1",
"created": "Sun, 23 Apr 2023 12:00:36 GMT"
}
] | 2023-04-25T00:00:00 | [
[
"Chen",
"Tinghui",
""
],
[
"Li",
"Shuai",
""
]
] | TITLE: Using Alternation Direction Method of Multipliers to Enhance robots
Calibration Accuracy based on Multi-Planal Constraints
ABSTRACT: With the widespread application of industrial robots, the problem of absolute
positioning accuracy becomes increasingly prominent. To ensure the working
state of the robots, researchers commonly adopt calibration techniques to
improve its accuracy. However, an industrial robot's working space is mostly
restricted in real working environments, making the collected samples fail in
covering the actual working space to result in the overall migration data. To
address this vital issue, this work proposes a novel industrial robot
calibrator that integrates a measurement configurations selection (MCS) method
and an alternation-direction-method-of-multipliers with multiple planes
constraints (AMPC) algorithm into its working process, whose ideas are
three-fold: a) selecting a group of optimal measurement configurations based on
the observability index to suppress the measurement noises, b) developing an
AMPC algorithm that evidently enhances the calibration accuracy and suppresses
the long-tail convergence, and c) proposing an industrial robot calibration
algorithm that incorporates MCS and AMPC to optimize an industrial robot's
kinematic parameters efficiently. For validating its performance, a
public-available dataset (HRS-P) is established on an HRS-JR680 industrial
robot. Extensive experimental results demonstrate that the proposed calibrator
outperforms several state-of-the-art models in calibration accuracy.
| new_dataset | 0.717563 |
2304.11631 | Dongjingdian Liu | Dongjingdin Liu, Pengpeng Chen, Miao Yao, Yijing Lu, Zijie Cai, Yuxin
Tian | TSGCNeXt: Dynamic-Static Multi-Graph Convolution for Efficient
Skeleton-Based Action Recognition with Long-term Learning Potential | null | null | null | null | cs.CV cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Skeleton-based action recognition has achieved remarkable results in human
action recognition with the development of graph convolutional networks (GCNs).
However, the recent works tend to construct complex learning mechanisms with
redundant training and exist a bottleneck for long time-series. To solve these
problems, we propose the Temporal-Spatio Graph ConvNeXt (TSGCNeXt) to explore
efficient learning mechanism of long temporal skeleton sequences. Firstly, a
new graph learning mechanism with simple structure, Dynamic-Static Separate
Multi-graph Convolution (DS-SMG) is proposed to aggregate features of multiple
independent topological graphs and avoid the node information being ignored
during dynamic convolution. Next, we construct a graph convolution training
acceleration mechanism to optimize the back-propagation computing of dynamic
graph learning with 55.08\% speed-up. Finally, the TSGCNeXt restructure the
overall structure of GCN with three Spatio-temporal learning
modules,efficiently modeling long temporal features. In comparison with
existing previous methods on large-scale datasets NTU RGB+D 60 and 120,
TSGCNeXt outperforms on single-stream networks. In addition, with the ema model
introduced into the multi-stream fusion, TSGCNeXt achieves SOTA levels. On the
cross-subject and cross-set of the NTU 120, accuracies reach 90.22% and 91.74%.
| [
{
"version": "v1",
"created": "Sun, 23 Apr 2023 12:10:36 GMT"
}
] | 2023-04-25T00:00:00 | [
[
"Liu",
"Dongjingdin",
""
],
[
"Chen",
"Pengpeng",
""
],
[
"Yao",
"Miao",
""
],
[
"Lu",
"Yijing",
""
],
[
"Cai",
"Zijie",
""
],
[
"Tian",
"Yuxin",
""
]
] | TITLE: TSGCNeXt: Dynamic-Static Multi-Graph Convolution for Efficient
Skeleton-Based Action Recognition with Long-term Learning Potential
ABSTRACT: Skeleton-based action recognition has achieved remarkable results in human
action recognition with the development of graph convolutional networks (GCNs).
However, the recent works tend to construct complex learning mechanisms with
redundant training and exist a bottleneck for long time-series. To solve these
problems, we propose the Temporal-Spatio Graph ConvNeXt (TSGCNeXt) to explore
efficient learning mechanism of long temporal skeleton sequences. Firstly, a
new graph learning mechanism with simple structure, Dynamic-Static Separate
Multi-graph Convolution (DS-SMG) is proposed to aggregate features of multiple
independent topological graphs and avoid the node information being ignored
during dynamic convolution. Next, we construct a graph convolution training
acceleration mechanism to optimize the back-propagation computing of dynamic
graph learning with 55.08\% speed-up. Finally, the TSGCNeXt restructure the
overall structure of GCN with three Spatio-temporal learning
modules,efficiently modeling long temporal features. In comparison with
existing previous methods on large-scale datasets NTU RGB+D 60 and 120,
TSGCNeXt outperforms on single-stream networks. In addition, with the ema model
introduced into the multi-stream fusion, TSGCNeXt achieves SOTA levels. On the
cross-subject and cross-set of the NTU 120, accuracies reach 90.22% and 91.74%.
| no_new_dataset | 0.71123 |
2304.11633 | Bo Li | Bo Li, Gexiang Fang, Yang Yang, Quansen Wang, Wei Ye, Wen Zhao, Shikun
Zhang | Evaluating ChatGPT's Information Extraction Capabilities: An Assessment
of Performance, Explainability, Calibration, and Faithfulness | null | null | null | null | cs.CL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The capability of Large Language Models (LLMs) like ChatGPT to comprehend
user intent and provide reasonable responses has made them extremely popular
lately. In this paper, we focus on assessing the overall ability of ChatGPT
using 7 fine-grained information extraction (IE) tasks. Specially, we present
the systematically analysis by measuring ChatGPT's performance, explainability,
calibration, and faithfulness, and resulting in 15 keys from either the ChatGPT
or domain experts. Our findings reveal that ChatGPT's performance in
Standard-IE setting is poor, but it surprisingly exhibits excellent performance
in the OpenIE setting, as evidenced by human evaluation. In addition, our
research indicates that ChatGPT provides high-quality and trustworthy
explanations for its decisions. However, there is an issue of ChatGPT being
overconfident in its predictions, which resulting in low calibration.
Furthermore, ChatGPT demonstrates a high level of faithfulness to the original
text in the majority of cases. We manually annotate and release the test sets
of 7 fine-grained IE tasks contains 14 datasets to further promote the
research. The datasets and code are available at
https://github.com/pkuserc/ChatGPT_for_IE.
| [
{
"version": "v1",
"created": "Sun, 23 Apr 2023 12:33:18 GMT"
}
] | 2023-04-25T00:00:00 | [
[
"Li",
"Bo",
""
],
[
"Fang",
"Gexiang",
""
],
[
"Yang",
"Yang",
""
],
[
"Wang",
"Quansen",
""
],
[
"Ye",
"Wei",
""
],
[
"Zhao",
"Wen",
""
],
[
"Zhang",
"Shikun",
""
]
] | TITLE: Evaluating ChatGPT's Information Extraction Capabilities: An Assessment
of Performance, Explainability, Calibration, and Faithfulness
ABSTRACT: The capability of Large Language Models (LLMs) like ChatGPT to comprehend
user intent and provide reasonable responses has made them extremely popular
lately. In this paper, we focus on assessing the overall ability of ChatGPT
using 7 fine-grained information extraction (IE) tasks. Specially, we present
the systematically analysis by measuring ChatGPT's performance, explainability,
calibration, and faithfulness, and resulting in 15 keys from either the ChatGPT
or domain experts. Our findings reveal that ChatGPT's performance in
Standard-IE setting is poor, but it surprisingly exhibits excellent performance
in the OpenIE setting, as evidenced by human evaluation. In addition, our
research indicates that ChatGPT provides high-quality and trustworthy
explanations for its decisions. However, there is an issue of ChatGPT being
overconfident in its predictions, which resulting in low calibration.
Furthermore, ChatGPT demonstrates a high level of faithfulness to the original
text in the majority of cases. We manually annotate and release the test sets
of 7 fine-grained IE tasks contains 14 datasets to further promote the
research. The datasets and code are available at
https://github.com/pkuserc/ChatGPT_for_IE.
| new_dataset | 0.706152 |
2304.11651 | Christin Katharina Kreutz | Christin Katharina Kreutz, Martin Blum, Philipp Schaer, Ralf Schenkel,
Benjamin Weyers | Evaluating Digital Library Search Systems by using Formal Process
Modelling | 10 pages + references, publication accepted at JCDL'23 | null | null | null | cs.DL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Evaluations of digital library information systems are typically centred on
users correctly, efficiently, and quickly performing predefined tasks.
Additionally, users generally enjoy working with the evaluated system, and
completed questionnaires show an interface's excellent user experience.
However, such evaluations do not explicitly consider comparing or connecting
user-specific information-seeking behaviour with digital library system
capabilities and thus overlook actual user needs or further system
requirements. We aim to close this gap by introducing the usage of
formalisations of users' task conduction strategies to compare their
information needs with the capabilities of such information systems. We observe
users' strategies in scope of expert finding and paper search. We propose and
investigate using the business process model notation to formalise task
conduction strategies and the SchenQL digital library interface as an example
system. We conduct interviews in a qualitative evaluation with 13 participants
from various backgrounds from which we derive models. We discovered that the
formalisations are suitable and helpful to mirror the strategies back to users
and to compare users' ideal task conductions with capabilities of information
systems. We conclude using formal models for qualitative digital library
studies being a suitable mean to identify current limitations and depict users'
task conduction strategies. Our published dataset containing the evaluation
data can be reused to investigate other digital library systems' fit for
depicting users' ideal task solutions.
| [
{
"version": "v1",
"created": "Sun, 23 Apr 2023 13:38:19 GMT"
}
] | 2023-04-25T00:00:00 | [
[
"Kreutz",
"Christin Katharina",
""
],
[
"Blum",
"Martin",
""
],
[
"Schaer",
"Philipp",
""
],
[
"Schenkel",
"Ralf",
""
],
[
"Weyers",
"Benjamin",
""
]
] | TITLE: Evaluating Digital Library Search Systems by using Formal Process
Modelling
ABSTRACT: Evaluations of digital library information systems are typically centred on
users correctly, efficiently, and quickly performing predefined tasks.
Additionally, users generally enjoy working with the evaluated system, and
completed questionnaires show an interface's excellent user experience.
However, such evaluations do not explicitly consider comparing or connecting
user-specific information-seeking behaviour with digital library system
capabilities and thus overlook actual user needs or further system
requirements. We aim to close this gap by introducing the usage of
formalisations of users' task conduction strategies to compare their
information needs with the capabilities of such information systems. We observe
users' strategies in scope of expert finding and paper search. We propose and
investigate using the business process model notation to formalise task
conduction strategies and the SchenQL digital library interface as an example
system. We conduct interviews in a qualitative evaluation with 13 participants
from various backgrounds from which we derive models. We discovered that the
formalisations are suitable and helpful to mirror the strategies back to users
and to compare users' ideal task conductions with capabilities of information
systems. We conclude using formal models for qualitative digital library
studies being a suitable mean to identify current limitations and depict users'
task conduction strategies. Our published dataset containing the evaluation
data can be reused to investigate other digital library systems' fit for
depicting users' ideal task solutions.
| new_dataset | 0.717377 |
2304.11658 | Lin Shu | Lin Shu, Chuan Chen, Zibin Zheng | Capturing Fine-grained Semantics in Contrastive Graph Representation
Learning | null | null | null | null | cs.LG | http://creativecommons.org/licenses/by/4.0/ | Graph contrastive learning defines a contrastive task to pull similar
instances close and push dissimilar instances away. It learns discriminative
node embeddings without supervised labels, which has aroused increasing
attention in the past few years. Nevertheless, existing methods of graph
contrastive learning ignore the differences between diverse semantics existed
in graphs, which learn coarse-grained node embeddings and lead to sub-optimal
performances on downstream tasks. To bridge this gap, we propose a novel
Fine-grained Semantics enhanced Graph Contrastive Learning (FSGCL) in this
paper. Concretely, FSGCL first introduces a motif-based graph construction,
which employs graph motifs to extract diverse semantics existed in graphs from
the perspective of input data. Then, the semantic-level contrastive task is
explored to further enhance the utilization of fine-grained semantics from the
perspective of model training. Experiments on five real-world datasets
demonstrate the superiority of our proposed FSGCL over state-of-the-art
methods. To make the results reproducible, we will make our codes public on
GitHub after this paper is accepted.
| [
{
"version": "v1",
"created": "Sun, 23 Apr 2023 14:05:05 GMT"
}
] | 2023-04-25T00:00:00 | [
[
"Shu",
"Lin",
""
],
[
"Chen",
"Chuan",
""
],
[
"Zheng",
"Zibin",
""
]
] | TITLE: Capturing Fine-grained Semantics in Contrastive Graph Representation
Learning
ABSTRACT: Graph contrastive learning defines a contrastive task to pull similar
instances close and push dissimilar instances away. It learns discriminative
node embeddings without supervised labels, which has aroused increasing
attention in the past few years. Nevertheless, existing methods of graph
contrastive learning ignore the differences between diverse semantics existed
in graphs, which learn coarse-grained node embeddings and lead to sub-optimal
performances on downstream tasks. To bridge this gap, we propose a novel
Fine-grained Semantics enhanced Graph Contrastive Learning (FSGCL) in this
paper. Concretely, FSGCL first introduces a motif-based graph construction,
which employs graph motifs to extract diverse semantics existed in graphs from
the perspective of input data. Then, the semantic-level contrastive task is
explored to further enhance the utilization of fine-grained semantics from the
perspective of model training. Experiments on five real-world datasets
demonstrate the superiority of our proposed FSGCL over state-of-the-art
methods. To make the results reproducible, we will make our codes public on
GitHub after this paper is accepted.
| no_new_dataset | 0.709025 |
2304.11662 | Hongyu Sun | Hongyu Sun, Yongcai Wang, Xudong Cai, Peng Wang, Zhe Huang, Deying Li,
Yu Shao, Shuo Wang | AirBirds: A Large-scale Challenging Dataset for Bird Strike Prevention
in Real-world Airports | 17 pages, 9 figures, 3 tables; accepted by ACCV 2022 | null | null | null | cs.CV | http://creativecommons.org/licenses/by-nc-sa/4.0/ | One fundamental limitation to the research of bird strike prevention is the
lack of a large-scale dataset taken directly from real-world airports. Existing
relevant datasets are either small in size or not dedicated for this purpose.
To advance the research and practical solutions for bird strike prevention, in
this paper, we present a large-scale challenging dataset AirBirds that consists
of 118,312 time-series images, where a total of 409,967 bounding boxes of
flying birds are manually, carefully annotated. The average size of all
annotated instances is smaller than 10 pixels in 1920x1080 images. Images in
the dataset are captured over 4 seasons of a whole year by a network of cameras
deployed at a real-world airport, covering diverse bird species, lighting
conditions and 13 meteorological scenarios. To the best of our knowledge, it is
the first large-scale image dataset that directly collects flying birds in
real-world airports for bird strike prevention. This dataset is publicly
available at https://airbirdsdata.github.io/.
| [
{
"version": "v1",
"created": "Sun, 23 Apr 2023 14:19:28 GMT"
}
] | 2023-04-25T00:00:00 | [
[
"Sun",
"Hongyu",
""
],
[
"Wang",
"Yongcai",
""
],
[
"Cai",
"Xudong",
""
],
[
"Wang",
"Peng",
""
],
[
"Huang",
"Zhe",
""
],
[
"Li",
"Deying",
""
],
[
"Shao",
"Yu",
""
],
[
"Wang",
"Shuo",
""
]
] | TITLE: AirBirds: A Large-scale Challenging Dataset for Bird Strike Prevention
in Real-world Airports
ABSTRACT: One fundamental limitation to the research of bird strike prevention is the
lack of a large-scale dataset taken directly from real-world airports. Existing
relevant datasets are either small in size or not dedicated for this purpose.
To advance the research and practical solutions for bird strike prevention, in
this paper, we present a large-scale challenging dataset AirBirds that consists
of 118,312 time-series images, where a total of 409,967 bounding boxes of
flying birds are manually, carefully annotated. The average size of all
annotated instances is smaller than 10 pixels in 1920x1080 images. Images in
the dataset are captured over 4 seasons of a whole year by a network of cameras
deployed at a real-world airport, covering diverse bird species, lighting
conditions and 13 meteorological scenarios. To the best of our knowledge, it is
the first large-scale image dataset that directly collects flying birds in
real-world airports for bird strike prevention. This dataset is publicly
available at https://airbirdsdata.github.io/.
| new_dataset | 0.715834 |
2304.11664 | Arash Ghafouri | Arash Ghafouri, Hasan Naderi, Mohammad Aghajani asl and Mahdi
Firouzmandi | IslamicPCQA: A Dataset for Persian Multi-hop Complex Question Answering
in Islamic Text Resources | null | null | null | null | cs.CL cs.AI | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Nowadays, one of the main challenges for Question Answering Systems is to
answer complex questions using various sources of information. Multi-hop
questions are a type of complex questions that require multi-step reasoning to
answer. In this article, the IslamicPCQA dataset is introduced. This is the
first Persian dataset for answering complex questions based on non-structured
information sources and consists of 12,282 question-answer pairs extracted from
9 Islamic encyclopedias. This dataset has been created inspired by the HotpotQA
English dataset approach, which was customized to suit the complexities of the
Persian language. Answering questions in this dataset requires more than one
paragraph and reasoning. The questions are not limited to any prior knowledge
base or ontology, and to provide robust reasoning ability, the dataset also
includes supporting facts and key sentences. The prepared dataset covers a wide
range of Islamic topics and aims to facilitate answering complex Persian
questions within this subject matter
| [
{
"version": "v1",
"created": "Sun, 23 Apr 2023 14:20:58 GMT"
}
] | 2023-04-25T00:00:00 | [
[
"Ghafouri",
"Arash",
""
],
[
"Naderi",
"Hasan",
""
],
[
"asl",
"Mohammad Aghajani",
""
],
[
"Firouzmandi",
"Mahdi",
""
]
] | TITLE: IslamicPCQA: A Dataset for Persian Multi-hop Complex Question Answering
in Islamic Text Resources
ABSTRACT: Nowadays, one of the main challenges for Question Answering Systems is to
answer complex questions using various sources of information. Multi-hop
questions are a type of complex questions that require multi-step reasoning to
answer. In this article, the IslamicPCQA dataset is introduced. This is the
first Persian dataset for answering complex questions based on non-structured
information sources and consists of 12,282 question-answer pairs extracted from
9 Islamic encyclopedias. This dataset has been created inspired by the HotpotQA
English dataset approach, which was customized to suit the complexities of the
Persian language. Answering questions in this dataset requires more than one
paragraph and reasoning. The questions are not limited to any prior knowledge
base or ontology, and to provide robust reasoning ability, the dataset also
includes supporting facts and key sentences. The prepared dataset covers a wide
range of Islamic topics and aims to facilitate answering complex Persian
questions within this subject matter
| new_dataset | 0.71566 |
2304.11665 | Zebang Shen | Zebang Shen, Hui Qian, Tongzhou Mu, Chao Zhang | Accelerated Doubly Stochastic Gradient Algorithm for Large-scale
Empirical Risk Minimization | Accepted to IJCAI 2017. Corresponding author: Hui Qian | null | null | null | cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Nowadays, algorithms with fast convergence, small memory footprints, and low
per-iteration complexity are particularly favorable for artificial intelligence
applications. In this paper, we propose a doubly stochastic algorithm with a
novel accelerating multi-momentum technique to solve large scale empirical risk
minimization problem for learning tasks. While enjoying a provably superior
convergence rate, in each iteration, such algorithm only accesses a mini batch
of samples and meanwhile updates a small block of variable coordinates, which
substantially reduces the amount of memory reference when both the massive
sample size and ultra-high dimensionality are involved. Empirical studies on
huge scale datasets are conducted to illustrate the efficiency of our method in
practice.
| [
{
"version": "v1",
"created": "Sun, 23 Apr 2023 14:21:29 GMT"
}
] | 2023-04-25T00:00:00 | [
[
"Shen",
"Zebang",
""
],
[
"Qian",
"Hui",
""
],
[
"Mu",
"Tongzhou",
""
],
[
"Zhang",
"Chao",
""
]
] | TITLE: Accelerated Doubly Stochastic Gradient Algorithm for Large-scale
Empirical Risk Minimization
ABSTRACT: Nowadays, algorithms with fast convergence, small memory footprints, and low
per-iteration complexity are particularly favorable for artificial intelligence
applications. In this paper, we propose a doubly stochastic algorithm with a
novel accelerating multi-momentum technique to solve large scale empirical risk
minimization problem for learning tasks. While enjoying a provably superior
convergence rate, in each iteration, such algorithm only accesses a mini batch
of samples and meanwhile updates a small block of variable coordinates, which
substantially reduces the amount of memory reference when both the massive
sample size and ultra-high dimensionality are involved. Empirical studies on
huge scale datasets are conducted to illustrate the efficiency of our method in
practice.
| no_new_dataset | 0.710603 |
2304.11670 | Qing Guo | Yang Hou, Qing Guo, Yihao Huang, Xiaofei Xie, Lei Ma, Jianjun Zhao | Evading DeepFake Detectors via Adversarial Statistical Consistency | Accepted by CVPR 2023 | null | null | null | cs.CV eess.IV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In recent years, as various realistic face forgery techniques known as
DeepFake improves by leaps and bounds,more and more DeepFake detection
techniques have been proposed. These methods typically rely on detecting
statistical differences between natural (i.e., real) and DeepFakegenerated
images in both spatial and frequency domains. In this work, we propose to
explicitly minimize the statistical differences to evade state-of-the-art
DeepFake detectors. To this end, we propose a statistical consistency attack
(StatAttack) against DeepFake detectors, which contains two main parts. First,
we select several statistical-sensitive natural degradations (i.e., exposure,
blur, and noise) and add them to the fake images in an adversarial way. Second,
we find that the statistical differences between natural and DeepFake images
are positively associated with the distribution shifting between the two kinds
of images, and we propose to use a distribution-aware loss to guide the
optimization of different degradations. As a result, the feature distributions
of generated adversarial examples is close to the natural images.Furthermore,
we extend the StatAttack to a more powerful version, MStatAttack, where we
extend the single-layer degradation to multi-layer degradations sequentially
and use the loss to tune the combination weights jointly. Comprehensive
experimental results on four spatial-based detectors and two frequency-based
detectors with four datasets demonstrate the effectiveness of our proposed
attack method in both white-box and black-box settings.
| [
{
"version": "v1",
"created": "Sun, 23 Apr 2023 14:40:42 GMT"
}
] | 2023-04-25T00:00:00 | [
[
"Hou",
"Yang",
""
],
[
"Guo",
"Qing",
""
],
[
"Huang",
"Yihao",
""
],
[
"Xie",
"Xiaofei",
""
],
[
"Ma",
"Lei",
""
],
[
"Zhao",
"Jianjun",
""
]
] | TITLE: Evading DeepFake Detectors via Adversarial Statistical Consistency
ABSTRACT: In recent years, as various realistic face forgery techniques known as
DeepFake improves by leaps and bounds,more and more DeepFake detection
techniques have been proposed. These methods typically rely on detecting
statistical differences between natural (i.e., real) and DeepFakegenerated
images in both spatial and frequency domains. In this work, we propose to
explicitly minimize the statistical differences to evade state-of-the-art
DeepFake detectors. To this end, we propose a statistical consistency attack
(StatAttack) against DeepFake detectors, which contains two main parts. First,
we select several statistical-sensitive natural degradations (i.e., exposure,
blur, and noise) and add them to the fake images in an adversarial way. Second,
we find that the statistical differences between natural and DeepFake images
are positively associated with the distribution shifting between the two kinds
of images, and we propose to use a distribution-aware loss to guide the
optimization of different degradations. As a result, the feature distributions
of generated adversarial examples is close to the natural images.Furthermore,
we extend the StatAttack to a more powerful version, MStatAttack, where we
extend the single-layer degradation to multi-layer degradations sequentially
and use the loss to tune the combination weights jointly. Comprehensive
experimental results on four spatial-based detectors and two frequency-based
detectors with four datasets demonstrate the effectiveness of our proposed
attack method in both white-box and black-box settings.
| no_new_dataset | 0.709252 |
2304.11681 | Ian Gray | Ian W. Gray, Jack Cable, Benjamin Brown, Vlad Cuiujuclu, Damon McCoy | Money Over Morals: A Business Analysis of Conti Ransomware | To be published in 2022 APWG Symposium on Electronic Crime Research
(eCrime) | null | null | null | cs.CR | http://creativecommons.org/licenses/by/4.0/ | Ransomware operations have evolved from relatively unsophisticated threat
actors into highly coordinated cybercrime syndicates that regularly extort
millions of dollars in a single attack. Despite dominating headlines and
crippling businesses across the globe, there is relatively little in-depth
research into the modern structure and economics of ransomware operations.
In this paper, we leverage leaked chat messages to provide an in-depth
empirical analysis of Conti, one of the largest ransomware groups. By analyzing
these chat messages, we construct a picture of Conti's operations as a
highly-profitable business, from profit structures to employee recruitment and
roles. We present novel methodologies to trace ransom payments, identifying
over $80 million in likely ransom payments to Conti and its predecessor -- over
five times as much as in previous public datasets. As part of our work, we
publish a dataset of 666 labeled Bitcoin addresses related to Conti and an
additional 75 Bitcoin addresses of likely ransom payments. Future work can
leverage this case study to more effectively trace -- and ultimately counteract
-- ransomware activity.
| [
{
"version": "v1",
"created": "Sun, 23 Apr 2023 15:21:38 GMT"
}
] | 2023-04-25T00:00:00 | [
[
"Gray",
"Ian W.",
""
],
[
"Cable",
"Jack",
""
],
[
"Brown",
"Benjamin",
""
],
[
"Cuiujuclu",
"Vlad",
""
],
[
"McCoy",
"Damon",
""
]
] | TITLE: Money Over Morals: A Business Analysis of Conti Ransomware
ABSTRACT: Ransomware operations have evolved from relatively unsophisticated threat
actors into highly coordinated cybercrime syndicates that regularly extort
millions of dollars in a single attack. Despite dominating headlines and
crippling businesses across the globe, there is relatively little in-depth
research into the modern structure and economics of ransomware operations.
In this paper, we leverage leaked chat messages to provide an in-depth
empirical analysis of Conti, one of the largest ransomware groups. By analyzing
these chat messages, we construct a picture of Conti's operations as a
highly-profitable business, from profit structures to employee recruitment and
roles. We present novel methodologies to trace ransom payments, identifying
over $80 million in likely ransom payments to Conti and its predecessor -- over
five times as much as in previous public datasets. As part of our work, we
publish a dataset of 666 labeled Bitcoin addresses related to Conti and an
additional 75 Bitcoin addresses of likely ransom payments. Future work can
leverage this case study to more effectively trace -- and ultimately counteract
-- ransomware activity.
| new_dataset | 0.714416 |
2304.11685 | Mathias Ibsen | Magnus Falkenberg, Anders Bensen Ottsen, Mathias Ibsen, Christian
Rathgeb | Child Face Recognition at Scale: Synthetic Data Generation and
Performance Benchmark | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We address the need for a large-scale database of children's faces by using
generative adversarial networks (GANs) and face age progression (FAP) models to
synthesize a realistic dataset referred to as HDA-SynChildFaces. To this end,
we proposed a processing pipeline that initially utilizes StyleGAN3 to sample
adult subjects, which are subsequently progressed to children of varying ages
using InterFaceGAN. Intra-subject variations, such as facial expression and
pose, are created by further manipulating the subjects in their latent space.
Additionally, the presented pipeline allows to evenly distribute the races of
subjects, allowing to generate a balanced and fair dataset with respect to race
distribution. The created HDA-SynChildFaces consists of 1,652 subjects and a
total of 188,832 images, each subject being present at various ages and with
many different intra-subject variations. Subsequently, we evaluates the
performance of various facial recognition systems on the generated database and
compare the results of adults and children at different ages. The study reveals
that children consistently perform worse than adults, on all tested systems,
and the degradation in performance is proportional to age. Additionally, our
study uncovers some biases in the recognition systems, with Asian and Black
subjects and females performing worse than White and Latino Hispanic subjects
and males.
| [
{
"version": "v1",
"created": "Sun, 23 Apr 2023 15:29:26 GMT"
}
] | 2023-04-25T00:00:00 | [
[
"Falkenberg",
"Magnus",
""
],
[
"Ottsen",
"Anders Bensen",
""
],
[
"Ibsen",
"Mathias",
""
],
[
"Rathgeb",
"Christian",
""
]
] | TITLE: Child Face Recognition at Scale: Synthetic Data Generation and
Performance Benchmark
ABSTRACT: We address the need for a large-scale database of children's faces by using
generative adversarial networks (GANs) and face age progression (FAP) models to
synthesize a realistic dataset referred to as HDA-SynChildFaces. To this end,
we proposed a processing pipeline that initially utilizes StyleGAN3 to sample
adult subjects, which are subsequently progressed to children of varying ages
using InterFaceGAN. Intra-subject variations, such as facial expression and
pose, are created by further manipulating the subjects in their latent space.
Additionally, the presented pipeline allows to evenly distribute the races of
subjects, allowing to generate a balanced and fair dataset with respect to race
distribution. The created HDA-SynChildFaces consists of 1,652 subjects and a
total of 188,832 images, each subject being present at various ages and with
many different intra-subject variations. Subsequently, we evaluates the
performance of various facial recognition systems on the generated database and
compare the results of adults and children at different ages. The study reveals
that children consistently perform worse than adults, on all tested systems,
and the degradation in performance is proportional to age. Additionally, our
study uncovers some biases in the recognition systems, with Asian and Black
subjects and females performing worse than White and Latino Hispanic subjects
and males.
| new_dataset | 0.715206 |
2304.11688 | Wei Ju | Wei Ju, Xiao Luo, Meng Qu, Yifan Wang, Chong Chen, Minghua Deng,
Xian-Sheng Hua, Ming Zhang | TGNN: A Joint Semi-supervised Framework for Graph-level Classification | Accepted by Proceedings of the Thirty-First International Joint
Conference on Artificial Intelligence (IJCAI 2022) | null | null | null | cs.LG cs.AI cs.IR | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | This paper studies semi-supervised graph classification, a crucial task with
a wide range of applications in social network analysis and bioinformatics.
Recent works typically adopt graph neural networks to learn graph-level
representations for classification, failing to explicitly leverage features
derived from graph topology (e.g., paths). Moreover, when labeled data is
scarce, these methods are far from satisfactory due to their insufficient
topology exploration of unlabeled data. We address the challenge by proposing a
novel semi-supervised framework called Twin Graph Neural Network (TGNN). To
explore graph structural information from complementary views, our TGNN has a
message passing module and a graph kernel module. To fully utilize unlabeled
data, for each module, we calculate the similarity of each unlabeled graph to
other labeled graphs in the memory bank and our consistency loss encourages
consistency between two similarity distributions in different embedding spaces.
The two twin modules collaborate with each other by exchanging instance
similarity knowledge to fully explore the structure information of both labeled
and unlabeled data. We evaluate our TGNN on various public datasets and show
that it achieves strong performance.
| [
{
"version": "v1",
"created": "Sun, 23 Apr 2023 15:42:11 GMT"
}
] | 2023-04-25T00:00:00 | [
[
"Ju",
"Wei",
""
],
[
"Luo",
"Xiao",
""
],
[
"Qu",
"Meng",
""
],
[
"Wang",
"Yifan",
""
],
[
"Chen",
"Chong",
""
],
[
"Deng",
"Minghua",
""
],
[
"Hua",
"Xian-Sheng",
""
],
[
"Zhang",
"Ming",
""
]
] | TITLE: TGNN: A Joint Semi-supervised Framework for Graph-level Classification
ABSTRACT: This paper studies semi-supervised graph classification, a crucial task with
a wide range of applications in social network analysis and bioinformatics.
Recent works typically adopt graph neural networks to learn graph-level
representations for classification, failing to explicitly leverage features
derived from graph topology (e.g., paths). Moreover, when labeled data is
scarce, these methods are far from satisfactory due to their insufficient
topology exploration of unlabeled data. We address the challenge by proposing a
novel semi-supervised framework called Twin Graph Neural Network (TGNN). To
explore graph structural information from complementary views, our TGNN has a
message passing module and a graph kernel module. To fully utilize unlabeled
data, for each module, we calculate the similarity of each unlabeled graph to
other labeled graphs in the memory bank and our consistency loss encourages
consistency between two similarity distributions in different embedding spaces.
The two twin modules collaborate with each other by exchanging instance
similarity knowledge to fully explore the structure information of both labeled
and unlabeled data. We evaluate our TGNN on various public datasets and show
that it achieves strong performance.
| no_new_dataset | 0.710409 |
2304.11697 | Xinyu Zhang | Xinyu Zhang, Zhiwei Li, Zhenhong Zou, Xin Gao, Yijin Xiong, Dafeng
Jin, Jun Li, and Huaping Liu | Informative Data Selection with Uncertainty for Multi-modal Object
Detection | null | null | null | null | cs.CV eess.IV | http://creativecommons.org/licenses/by/4.0/ | Noise has always been nonnegligible trouble in object detection by creating
confusion in model reasoning, thereby reducing the informativeness of the data.
It can lead to inaccurate recognition due to the shift in the observed pattern,
that requires a robust generalization of the models. To implement a general
vision model, we need to develop deep learning models that can adaptively
select valid information from multi-modal data. This is mainly based on two
reasons. Multi-modal learning can break through the inherent defects of
single-modal data, and adaptive information selection can reduce chaos in
multi-modal data. To tackle this problem, we propose a universal
uncertainty-aware multi-modal fusion model. It adopts a multi-pipeline loosely
coupled architecture to combine the features and results from point clouds and
images. To quantify the correlation in multi-modal information, we model the
uncertainty, as the inverse of data information, in different modalities and
embed it in the bounding box generation. In this way, our model reduces the
randomness in fusion and generates reliable output. Moreover, we conducted a
completed investigation on the KITTI 2D object detection dataset and its
derived dirty data. Our fusion model is proven to resist severe noise
interference like Gaussian, motion blur, and frost, with only slight
degradation. The experiment results demonstrate the benefits of our adaptive
fusion. Our analysis on the robustness of multi-modal fusion will provide
further insights for future research.
| [
{
"version": "v1",
"created": "Sun, 23 Apr 2023 16:36:13 GMT"
}
] | 2023-04-25T00:00:00 | [
[
"Zhang",
"Xinyu",
""
],
[
"Li",
"Zhiwei",
""
],
[
"Zou",
"Zhenhong",
""
],
[
"Gao",
"Xin",
""
],
[
"Xiong",
"Yijin",
""
],
[
"Jin",
"Dafeng",
""
],
[
"Li",
"Jun",
""
],
[
"Liu",
"Huaping",
""
]
] | TITLE: Informative Data Selection with Uncertainty for Multi-modal Object
Detection
ABSTRACT: Noise has always been nonnegligible trouble in object detection by creating
confusion in model reasoning, thereby reducing the informativeness of the data.
It can lead to inaccurate recognition due to the shift in the observed pattern,
that requires a robust generalization of the models. To implement a general
vision model, we need to develop deep learning models that can adaptively
select valid information from multi-modal data. This is mainly based on two
reasons. Multi-modal learning can break through the inherent defects of
single-modal data, and adaptive information selection can reduce chaos in
multi-modal data. To tackle this problem, we propose a universal
uncertainty-aware multi-modal fusion model. It adopts a multi-pipeline loosely
coupled architecture to combine the features and results from point clouds and
images. To quantify the correlation in multi-modal information, we model the
uncertainty, as the inverse of data information, in different modalities and
embed it in the bounding box generation. In this way, our model reduces the
randomness in fusion and generates reliable output. Moreover, we conducted a
completed investigation on the KITTI 2D object detection dataset and its
derived dirty data. Our fusion model is proven to resist severe noise
interference like Gaussian, motion blur, and frost, with only slight
degradation. The experiment results demonstrate the benefits of our adaptive
fusion. Our analysis on the robustness of multi-modal fusion will provide
further insights for future research.
| no_new_dataset | 0.708231 |
2304.11701 | Di Wang | Di Wang, Bo Du, Liangpei Zhang, and Dacheng Tao | HKNAS: Classification of Hyperspectral Imagery Based on Hyper Kernel
Neural Architecture Search | Accepted by IEEE TNNLS. The code will be released at
https://github.com/DotWang/HKNAS | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Recent neural architecture search (NAS) based approaches have made great
progress in hyperspectral image (HSI) classification tasks. However, the
architectures are usually optimized independently of the network weights,
increasing searching time and restricting model performances. To tackle these
issues, in this paper, different from previous methods that extra define
structural parameters, we propose to directly generate structural parameters by
utilizing the specifically designed hyper kernels, ingeniously converting the
original complex dual optimization problem into easily implemented one-tier
optimizations, and greatly shrinking searching costs. Then, we develop a
hierarchical multi-module search space whose candidate operations only contain
convolutions, and these operations can be integrated into unified kernels.
Using the above searching strategy and searching space, we obtain three kinds
of networks to separately conduct pixel-level or image-level classifications
with 1-D or 3-D convolutions. In addition, by combining the proposed hyper
kernel searching scheme with the 3-D convolution decomposition mechanism, we
obtain diverse architectures to simulate 3-D convolutions, greatly improving
network flexibilities. A series of quantitative and qualitative experiments on
six public datasets demonstrate that the proposed methods achieve
state-of-the-art results compared with other advanced NAS-based HSI
classification approaches.
| [
{
"version": "v1",
"created": "Sun, 23 Apr 2023 17:27:40 GMT"
}
] | 2023-04-25T00:00:00 | [
[
"Wang",
"Di",
""
],
[
"Du",
"Bo",
""
],
[
"Zhang",
"Liangpei",
""
],
[
"Tao",
"Dacheng",
""
]
] | TITLE: HKNAS: Classification of Hyperspectral Imagery Based on Hyper Kernel
Neural Architecture Search
ABSTRACT: Recent neural architecture search (NAS) based approaches have made great
progress in hyperspectral image (HSI) classification tasks. However, the
architectures are usually optimized independently of the network weights,
increasing searching time and restricting model performances. To tackle these
issues, in this paper, different from previous methods that extra define
structural parameters, we propose to directly generate structural parameters by
utilizing the specifically designed hyper kernels, ingeniously converting the
original complex dual optimization problem into easily implemented one-tier
optimizations, and greatly shrinking searching costs. Then, we develop a
hierarchical multi-module search space whose candidate operations only contain
convolutions, and these operations can be integrated into unified kernels.
Using the above searching strategy and searching space, we obtain three kinds
of networks to separately conduct pixel-level or image-level classifications
with 1-D or 3-D convolutions. In addition, by combining the proposed hyper
kernel searching scheme with the 3-D convolution decomposition mechanism, we
obtain diverse architectures to simulate 3-D convolutions, greatly improving
network flexibilities. A series of quantitative and qualitative experiments on
six public datasets demonstrate that the proposed methods achieve
state-of-the-art results compared with other advanced NAS-based HSI
classification approaches.
| no_new_dataset | 0.710672 |
2304.11708 | Wonjun Yi | Wonjun Yi, Jung-Woo Choi and Jae-Woo Lee | Sound-based drone fault classification using multitask learning | Accepted at 29th International Congress on Sound and Vibration
(ICSV29). Dataset available: https://zenodo.org/record/7779574#.ZEVncnZBwQ- | null | null | null | cs.SD cs.AI eess.AS | http://creativecommons.org/licenses/by-nc-nd/4.0/ | The drone has been used for various purposes, including military
applications, aerial photography, and pesticide spraying. However, the drone is
vulnerable to external disturbances, and malfunction in propellers and motors
can easily occur. To improve the safety of drone operations, one should detect
the mechanical faults of drones in real-time. This paper proposes a sound-based
deep neural network (DNN) fault classifier and drone sound dataset. The dataset
was constructed by collecting the operating sounds of drones from microphones
mounted on three different drones in an anechoic chamber. The dataset includes
various operating conditions of drones, such as flight directions (front, back,
right, left, clockwise, counterclockwise) and faults on propellers and motors.
The drone sounds were then mixed with noises recorded in five different spots
on the university campus, with a signal-to-noise ratio (SNR) varying from 10 dB
to 15 dB. Using the acquired dataset, we train a DNN classifier, 1DCNN-ResNet,
that classifies the types of mechanical faults and their locations from
short-time input waveforms. We employ multitask learning (MTL) and incorporate
the direction classification task as an auxiliary task to make the classifier
learn more general audio features. The test over unseen data reveals that the
proposed multitask model can successfully classify faults in drones and
outperforms single-task models even with less training data.
| [
{
"version": "v1",
"created": "Sun, 23 Apr 2023 17:55:40 GMT"
}
] | 2023-04-25T00:00:00 | [
[
"Yi",
"Wonjun",
""
],
[
"Choi",
"Jung-Woo",
""
],
[
"Lee",
"Jae-Woo",
""
]
] | TITLE: Sound-based drone fault classification using multitask learning
ABSTRACT: The drone has been used for various purposes, including military
applications, aerial photography, and pesticide spraying. However, the drone is
vulnerable to external disturbances, and malfunction in propellers and motors
can easily occur. To improve the safety of drone operations, one should detect
the mechanical faults of drones in real-time. This paper proposes a sound-based
deep neural network (DNN) fault classifier and drone sound dataset. The dataset
was constructed by collecting the operating sounds of drones from microphones
mounted on three different drones in an anechoic chamber. The dataset includes
various operating conditions of drones, such as flight directions (front, back,
right, left, clockwise, counterclockwise) and faults on propellers and motors.
The drone sounds were then mixed with noises recorded in five different spots
on the university campus, with a signal-to-noise ratio (SNR) varying from 10 dB
to 15 dB. Using the acquired dataset, we train a DNN classifier, 1DCNN-ResNet,
that classifies the types of mechanical faults and their locations from
short-time input waveforms. We employ multitask learning (MTL) and incorporate
the direction classification task as an auxiliary task to make the classifier
learn more general audio features. The test over unseen data reveals that the
proposed multitask model can successfully classify faults in drones and
outperforms single-task models even with less training data.
| new_dataset | 0.72078 |
2304.11718 | Ihab Bendidi | Ihab Bendidi, Adrien Bardes, Ethan Cohen, Alexis Lamiable, Guillaume
Bollot, Auguste Genovesio | No Free Lunch in Self Supervised Representation Learning | null | null | null | null | cs.CV cs.AI | http://creativecommons.org/licenses/by/4.0/ | Self-supervised representation learning in computer vision relies heavily on
hand-crafted image transformations to learn meaningful and invariant features.
However few extensive explorations of the impact of transformation design have
been conducted in the literature. In particular, the dependence of downstream
performances to transformation design has been established, but not studied in
depth. In this work, we explore this relationship, its impact on a domain other
than natural images, and show that designing the transformations can be viewed
as a form of supervision. First, we demonstrate that not only do
transformations have an effect on downstream performance and relevance of
clustering, but also that each category in a supervised dataset can be impacted
in a different way. Following this, we explore the impact of transformation
design on microscopy images, a domain where the difference between classes is
more subtle and fuzzy than in natural images. In this case, we observe a
greater impact on downstream tasks performances. Finally, we demonstrate that
transformation design can be leveraged as a form of supervision, as careful
selection of these by a domain expert can lead to a drastic increase in
performance on a given downstream task.
| [
{
"version": "v1",
"created": "Sun, 23 Apr 2023 18:14:19 GMT"
}
] | 2023-04-25T00:00:00 | [
[
"Bendidi",
"Ihab",
""
],
[
"Bardes",
"Adrien",
""
],
[
"Cohen",
"Ethan",
""
],
[
"Lamiable",
"Alexis",
""
],
[
"Bollot",
"Guillaume",
""
],
[
"Genovesio",
"Auguste",
""
]
] | TITLE: No Free Lunch in Self Supervised Representation Learning
ABSTRACT: Self-supervised representation learning in computer vision relies heavily on
hand-crafted image transformations to learn meaningful and invariant features.
However few extensive explorations of the impact of transformation design have
been conducted in the literature. In particular, the dependence of downstream
performances to transformation design has been established, but not studied in
depth. In this work, we explore this relationship, its impact on a domain other
than natural images, and show that designing the transformations can be viewed
as a form of supervision. First, we demonstrate that not only do
transformations have an effect on downstream performance and relevance of
clustering, but also that each category in a supervised dataset can be impacted
in a different way. Following this, we explore the impact of transformation
design on microscopy images, a domain where the difference between classes is
more subtle and fuzzy than in natural images. In this case, we observe a
greater impact on downstream tasks performances. Finally, we demonstrate that
transformation design can be leveraged as a form of supervision, as careful
selection of these by a domain expert can lead to a drastic increase in
performance on a given downstream task.
| no_new_dataset | 0.711268 |
2304.11721 | Zhichao Xu | Zhichao Xu, Daniel Cohen | A Lightweight Constrained Generation Alternative for Query-focused
Summarization | in proceedings of SIGIR 2023 | null | null | null | cs.IR cs.CL | http://creativecommons.org/licenses/by/4.0/ | Query-focused summarization (QFS) aims to provide a summary of a document
that satisfies information need of a given query and is useful in various IR
applications, such as abstractive snippet generation. Current QFS approaches
typically involve injecting additional information, e.g. query-answer relevance
or fine-grained token-level interaction between a query and document, into a
finetuned large language model. However, these approaches often require extra
parameters \& training, and generalize poorly to new dataset distributions. To
mitigate this, we propose leveraging a recently developed constrained
generation model Neurological Decoding (NLD) as an alternative to current QFS
regimes which rely on additional sub-architectures and training. We first
construct lexical constraints by identifying important tokens from the document
using a lightweight gradient attribution model, then subsequently force the
generated summary to satisfy these constraints by directly manipulating the
final vocabulary likelihood. This lightweight approach requires no additional
parameters or finetuning as it utilizes both an off-the-shelf neural retrieval
model to construct the constraints and a standard generative language model to
produce the QFS. We demonstrate the efficacy of this approach on two public QFS
collections achieving near parity with the state-of-the-art model with
substantially reduced complexity.
| [
{
"version": "v1",
"created": "Sun, 23 Apr 2023 18:43:48 GMT"
}
] | 2023-04-25T00:00:00 | [
[
"Xu",
"Zhichao",
""
],
[
"Cohen",
"Daniel",
""
]
] | TITLE: A Lightweight Constrained Generation Alternative for Query-focused
Summarization
ABSTRACT: Query-focused summarization (QFS) aims to provide a summary of a document
that satisfies information need of a given query and is useful in various IR
applications, such as abstractive snippet generation. Current QFS approaches
typically involve injecting additional information, e.g. query-answer relevance
or fine-grained token-level interaction between a query and document, into a
finetuned large language model. However, these approaches often require extra
parameters \& training, and generalize poorly to new dataset distributions. To
mitigate this, we propose leveraging a recently developed constrained
generation model Neurological Decoding (NLD) as an alternative to current QFS
regimes which rely on additional sub-architectures and training. We first
construct lexical constraints by identifying important tokens from the document
using a lightweight gradient attribution model, then subsequently force the
generated summary to satisfy these constraints by directly manipulating the
final vocabulary likelihood. This lightweight approach requires no additional
parameters or finetuning as it utilizes both an off-the-shelf neural retrieval
model to construct the constraints and a standard generative language model to
produce the QFS. We demonstrate the efficacy of this approach on two public QFS
collections achieving near parity with the state-of-the-art model with
substantially reduced complexity.
| no_new_dataset | 0.712245 |
2304.11722 | Zhenwei Tang | Zhenwei Tang, Griffin Floto, Armin Toroghi, Shichao Pei, Xiangliang
Zhang, Scott Sanner | LogicRec: Recommendation with Users' Logical Requirements | SIGIR 2023 | null | 10.1145/3539618.3592012 | null | cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Users may demand recommendations with highly personalized requirements
involving logical operations, e.g., the intersection of two requirements, where
such requirements naturally form structured logical queries on knowledge graphs
(KGs). To date, existing recommender systems lack the capability to tackle
users' complex logical requirements. In this work, we formulate the problem of
recommendation with users' logical requirements (LogicRec) and construct
benchmark datasets for LogicRec. Furthermore, we propose an initial solution
for LogicRec based on logical requirement retrieval and user preference
retrieval, where we face two challenges. First, KGs are incomplete in nature.
Therefore, there are always missing true facts, which entails that the answers
to logical requirements can not be completely found in KGs. In this case, item
selection based on the answers to logical queries is not applicable. We thus
resort to logical query embedding (LQE) to jointly infer missing facts and
retrieve items based on logical requirements. Second, answer sets are
under-exploited. Existing LQE methods can only deal with query-answer pairs,
where queries in our case are the intersected user preferences and logical
requirements. However, the logical requirements and user preferences have
different answer sets, offering us richer knowledge about the requirements and
preferences by providing requirement-item and preference-item pairs. Thus, we
design a multi-task knowledge-sharing mechanism to exploit these answer sets
collectively. Extensive experimental results demonstrate the significance of
the LogicRec task and the effectiveness of our proposed method.
| [
{
"version": "v1",
"created": "Sun, 23 Apr 2023 18:46:58 GMT"
}
] | 2023-04-25T00:00:00 | [
[
"Tang",
"Zhenwei",
""
],
[
"Floto",
"Griffin",
""
],
[
"Toroghi",
"Armin",
""
],
[
"Pei",
"Shichao",
""
],
[
"Zhang",
"Xiangliang",
""
],
[
"Sanner",
"Scott",
""
]
] | TITLE: LogicRec: Recommendation with Users' Logical Requirements
ABSTRACT: Users may demand recommendations with highly personalized requirements
involving logical operations, e.g., the intersection of two requirements, where
such requirements naturally form structured logical queries on knowledge graphs
(KGs). To date, existing recommender systems lack the capability to tackle
users' complex logical requirements. In this work, we formulate the problem of
recommendation with users' logical requirements (LogicRec) and construct
benchmark datasets for LogicRec. Furthermore, we propose an initial solution
for LogicRec based on logical requirement retrieval and user preference
retrieval, where we face two challenges. First, KGs are incomplete in nature.
Therefore, there are always missing true facts, which entails that the answers
to logical requirements can not be completely found in KGs. In this case, item
selection based on the answers to logical queries is not applicable. We thus
resort to logical query embedding (LQE) to jointly infer missing facts and
retrieve items based on logical requirements. Second, answer sets are
under-exploited. Existing LQE methods can only deal with query-answer pairs,
where queries in our case are the intersected user preferences and logical
requirements. However, the logical requirements and user preferences have
different answer sets, offering us richer knowledge about the requirements and
preferences by providing requirement-item and preference-item pairs. Thus, we
design a multi-task knowledge-sharing mechanism to exploit these answer sets
collectively. Extensive experimental results demonstrate the significance of
the LogicRec task and the effectiveness of our proposed method.
| no_new_dataset | 0.704973 |
2304.11732 | Xiaozhe Yin | Xiaozhe Yin, Masoud Fallah-Shorshani, Rob McConnell, Scott Fruin,
Yao-Yi Chiang, Meredith Franklin | Quantile Extreme Gradient Boosting for Uncertainty Quantification | 15 pages, 7 figures and 4 tables | null | null | null | stat.ML cs.LG | http://creativecommons.org/licenses/by/4.0/ | As the availability, size and complexity of data have increased in recent
years, machine learning (ML) techniques have become popular for modeling.
Predictions resulting from applying ML models are often used for inference,
decision-making, and downstream applications. A crucial yet often overlooked
aspect of ML is uncertainty quantification, which can significantly impact how
predictions from models are used and interpreted.
Extreme Gradient Boosting (XGBoost) is one of the most popular ML methods
given its simple implementation, fast computation, and sequential learning,
which make its predictions highly accurate compared to other methods. However,
techniques for uncertainty determination in ML models such as XGBoost have not
yet been universally agreed among its varying applications. We propose
enhancements to XGBoost whereby a modified quantile regression is used as the
objective function to estimate uncertainty (QXGBoost). Specifically, we
included the Huber norm in the quantile regression model to construct a
differentiable approximation to the quantile regression error function. This
key step allows XGBoost, which uses a gradient-based optimization algorithm, to
make probabilistic predictions efficiently.
QXGBoost was applied to create 90\% prediction intervals for one simulated
dataset and one real-world environmental dataset of measured traffic noise. Our
proposed method had comparable or better performance than the uncertainty
estimates generated for regular and quantile light gradient boosting. For both
the simulated and traffic noise datasets, the overall performance of the
prediction intervals from QXGBoost were better than other models based on
coverage width-based criterion.
| [
{
"version": "v1",
"created": "Sun, 23 Apr 2023 19:46:19 GMT"
}
] | 2023-04-25T00:00:00 | [
[
"Yin",
"Xiaozhe",
""
],
[
"Fallah-Shorshani",
"Masoud",
""
],
[
"McConnell",
"Rob",
""
],
[
"Fruin",
"Scott",
""
],
[
"Chiang",
"Yao-Yi",
""
],
[
"Franklin",
"Meredith",
""
]
] | TITLE: Quantile Extreme Gradient Boosting for Uncertainty Quantification
ABSTRACT: As the availability, size and complexity of data have increased in recent
years, machine learning (ML) techniques have become popular for modeling.
Predictions resulting from applying ML models are often used for inference,
decision-making, and downstream applications. A crucial yet often overlooked
aspect of ML is uncertainty quantification, which can significantly impact how
predictions from models are used and interpreted.
Extreme Gradient Boosting (XGBoost) is one of the most popular ML methods
given its simple implementation, fast computation, and sequential learning,
which make its predictions highly accurate compared to other methods. However,
techniques for uncertainty determination in ML models such as XGBoost have not
yet been universally agreed among its varying applications. We propose
enhancements to XGBoost whereby a modified quantile regression is used as the
objective function to estimate uncertainty (QXGBoost). Specifically, we
included the Huber norm in the quantile regression model to construct a
differentiable approximation to the quantile regression error function. This
key step allows XGBoost, which uses a gradient-based optimization algorithm, to
make probabilistic predictions efficiently.
QXGBoost was applied to create 90\% prediction intervals for one simulated
dataset and one real-world environmental dataset of measured traffic noise. Our
proposed method had comparable or better performance than the uncertainty
estimates generated for regular and quantile light gradient boosting. For both
the simulated and traffic noise datasets, the overall performance of the
prediction intervals from QXGBoost were better than other models based on
coverage width-based criterion.
| no_new_dataset | 0.70318 |
2304.11734 | Mehdi Rafiei | Mehdi Rafiei and Alexandros Iosifidis | Class-Specific Variational Auto-Encoder for Content-Based Image
Retrieval | 8 pages, 7 figures, 6 tables, accepted at IJCNN conference | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Using a discriminative representation obtained by supervised deep learning
methods showed promising results on diverse Content-Based Image Retrieval
(CBIR) problems. However, existing methods exploiting labels during training
try to discriminate all available classes, which is not ideal in cases where
the retrieval problem focuses on a class of interest. In this paper, we propose
a regularized loss for Variational Auto-Encoders (VAEs) forcing the model to
focus on a given class of interest. As a result, the model learns to
discriminate the data belonging to the class of interest from any other
possibility, making the learnt latent space of the VAE suitable for
class-specific retrieval tasks. The proposed Class-Specific Variational
Auto-Encoder (CS-VAE) is evaluated on three public and one custom datasets, and
its performance is compared with that of three related VAE-based methods.
Experimental results show that the proposed method outperforms its competition
in both in-domain and out-of-domain retrieval problems.
| [
{
"version": "v1",
"created": "Sun, 23 Apr 2023 19:51:25 GMT"
}
] | 2023-04-25T00:00:00 | [
[
"Rafiei",
"Mehdi",
""
],
[
"Iosifidis",
"Alexandros",
""
]
] | TITLE: Class-Specific Variational Auto-Encoder for Content-Based Image
Retrieval
ABSTRACT: Using a discriminative representation obtained by supervised deep learning
methods showed promising results on diverse Content-Based Image Retrieval
(CBIR) problems. However, existing methods exploiting labels during training
try to discriminate all available classes, which is not ideal in cases where
the retrieval problem focuses on a class of interest. In this paper, we propose
a regularized loss for Variational Auto-Encoders (VAEs) forcing the model to
focus on a given class of interest. As a result, the model learns to
discriminate the data belonging to the class of interest from any other
possibility, making the learnt latent space of the VAE suitable for
class-specific retrieval tasks. The proposed Class-Specific Variational
Auto-Encoder (CS-VAE) is evaluated on three public and one custom datasets, and
its performance is compared with that of three related VAE-based methods.
Experimental results show that the proposed method outperforms its competition
in both in-domain and out-of-domain retrieval problems.
| no_new_dataset | 0.710377 |
2304.11740 | Sariah Mghames Dr | Sariah Mghames, Luca Castri, Marc Hanheide, Nicola Bellotto | A Neuro-Symbolic Approach for Enhanced Human Motion Prediction | International Joint Conference on Neural Networks (IJCNN), 2023 | null | null | null | cs.AI | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Reasoning on the context of human beings is crucial for many real-world
applications especially for those deploying autonomous systems (e.g. robots).
In this paper, we present a new approach for context reasoning to further
advance the field of human motion prediction. We therefore propose a
neuro-symbolic approach for human motion prediction (NeuroSyM), which weights
differently the interactions in the neighbourhood by leveraging an intuitive
technique for spatial representation called Qualitative Trajectory Calculus
(QTC). The proposed approach is experimentally tested on medium and long term
time horizons using two architectures from the state of art, one of which is a
baseline for human motion prediction and the other is a baseline for generic
multivariate time-series prediction. Six datasets of challenging crowded
scenarios, collected from both fixed and mobile cameras, were used for testing.
Experimental results show that the NeuroSyM approach outperforms in most cases
the baseline architectures in terms of prediction accuracy.
| [
{
"version": "v1",
"created": "Sun, 23 Apr 2023 20:11:40 GMT"
}
] | 2023-04-25T00:00:00 | [
[
"Mghames",
"Sariah",
""
],
[
"Castri",
"Luca",
""
],
[
"Hanheide",
"Marc",
""
],
[
"Bellotto",
"Nicola",
""
]
] | TITLE: A Neuro-Symbolic Approach for Enhanced Human Motion Prediction
ABSTRACT: Reasoning on the context of human beings is crucial for many real-world
applications especially for those deploying autonomous systems (e.g. robots).
In this paper, we present a new approach for context reasoning to further
advance the field of human motion prediction. We therefore propose a
neuro-symbolic approach for human motion prediction (NeuroSyM), which weights
differently the interactions in the neighbourhood by leveraging an intuitive
technique for spatial representation called Qualitative Trajectory Calculus
(QTC). The proposed approach is experimentally tested on medium and long term
time horizons using two architectures from the state of art, one of which is a
baseline for human motion prediction and the other is a baseline for generic
multivariate time-series prediction. Six datasets of challenging crowded
scenarios, collected from both fixed and mobile cameras, were used for testing.
Experimental results show that the NeuroSyM approach outperforms in most cases
the baseline architectures in terms of prediction accuracy.
| no_new_dataset | 0.711387 |
2304.11743 | Hoang Le | Hoang M. Le, Brian Price, Scott Cohen, Michael S. Brown | GamutMLP: A Lightweight MLP for Color Loss Recovery | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Cameras and image-editing software often process images in the wide-gamut
ProPhoto color space, encompassing 90% of all visible colors. However, when
images are encoded for sharing, this color-rich representation is transformed
and clipped to fit within the small-gamut standard RGB (sRGB) color space,
representing only 30% of visible colors. Recovering the lost color information
is challenging due to the clipping procedure. Inspired by neural implicit
representations for 2D images, we propose a method that optimizes a lightweight
multi-layer-perceptron (MLP) model during the gamut reduction step to predict
the clipped values. GamutMLP takes approximately 2 seconds to optimize and
requires only 23 KB of storage. The small memory footprint allows our GamutMLP
model to be saved as metadata in the sRGB image -- the model can be extracted
when needed to restore wide-gamut color values. We demonstrate the
effectiveness of our approach for color recovery and compare it with
alternative strategies, including pre-trained DNN-based gamut expansion
networks and other implicit neural representation methods. As part of this
effort, we introduce a new color gamut dataset of 2200 wide-gamut/small-gamut
images for training and testing. Our code and dataset can be found on the
project website: https://gamut-mlp.github.io.
| [
{
"version": "v1",
"created": "Sun, 23 Apr 2023 20:26:11 GMT"
}
] | 2023-04-25T00:00:00 | [
[
"Le",
"Hoang M.",
""
],
[
"Price",
"Brian",
""
],
[
"Cohen",
"Scott",
""
],
[
"Brown",
"Michael S.",
""
]
] | TITLE: GamutMLP: A Lightweight MLP for Color Loss Recovery
ABSTRACT: Cameras and image-editing software often process images in the wide-gamut
ProPhoto color space, encompassing 90% of all visible colors. However, when
images are encoded for sharing, this color-rich representation is transformed
and clipped to fit within the small-gamut standard RGB (sRGB) color space,
representing only 30% of visible colors. Recovering the lost color information
is challenging due to the clipping procedure. Inspired by neural implicit
representations for 2D images, we propose a method that optimizes a lightweight
multi-layer-perceptron (MLP) model during the gamut reduction step to predict
the clipped values. GamutMLP takes approximately 2 seconds to optimize and
requires only 23 KB of storage. The small memory footprint allows our GamutMLP
model to be saved as metadata in the sRGB image -- the model can be extracted
when needed to restore wide-gamut color values. We demonstrate the
effectiveness of our approach for color recovery and compare it with
alternative strategies, including pre-trained DNN-based gamut expansion
networks and other implicit neural representation methods. As part of this
effort, we introduce a new color gamut dataset of 2200 wide-gamut/small-gamut
images for training and testing. Our code and dataset can be found on the
project website: https://gamut-mlp.github.io.
| new_dataset | 0.715623 |
2304.11749 | Zhi Chen | Zhi Chen, Sarah Tan, Urszula Chajewska, Cynthia Rudin, Rich Caruana | Missing Values and Imputation in Healthcare Data: Can Interpretable
Machine Learning Help? | Preprint of a paper accepted by CHIL 2023 | null | null | null | cs.LG | http://creativecommons.org/publicdomain/zero/1.0/ | Missing values are a fundamental problem in data science. Many datasets have
missing values that must be properly handled because the way missing values are
treated can have large impact on the resulting machine learning model. In
medical applications, the consequences may affect healthcare decisions. There
are many methods in the literature for dealing with missing values, including
state-of-the-art methods which often depend on black-box models for imputation.
In this work, we show how recent advances in interpretable machine learning
provide a new perspective for understanding and tackling the missing value
problem. We propose methods based on high-accuracy glass-box Explainable
Boosting Machines (EBMs) that can help users (1) gain new insights on
missingness mechanisms and better understand the causes of missingness, and (2)
detect -- or even alleviate -- potential risks introduced by imputation
algorithms. Experiments on real-world medical datasets illustrate the
effectiveness of the proposed methods.
| [
{
"version": "v1",
"created": "Sun, 23 Apr 2023 20:54:47 GMT"
}
] | 2023-04-25T00:00:00 | [
[
"Chen",
"Zhi",
""
],
[
"Tan",
"Sarah",
""
],
[
"Chajewska",
"Urszula",
""
],
[
"Rudin",
"Cynthia",
""
],
[
"Caruana",
"Rich",
""
]
] | TITLE: Missing Values and Imputation in Healthcare Data: Can Interpretable
Machine Learning Help?
ABSTRACT: Missing values are a fundamental problem in data science. Many datasets have
missing values that must be properly handled because the way missing values are
treated can have large impact on the resulting machine learning model. In
medical applications, the consequences may affect healthcare decisions. There
are many methods in the literature for dealing with missing values, including
state-of-the-art methods which often depend on black-box models for imputation.
In this work, we show how recent advances in interpretable machine learning
provide a new perspective for understanding and tackling the missing value
problem. We propose methods based on high-accuracy glass-box Explainable
Boosting Machines (EBMs) that can help users (1) gain new insights on
missingness mechanisms and better understand the causes of missingness, and (2)
detect -- or even alleviate -- potential risks introduced by imputation
algorithms. Experiments on real-world medical datasets illustrate the
effectiveness of the proposed methods.
| no_new_dataset | 0.710434 |
2304.11750 | Zhijun Liu | Zhijun Liu, Yiwei Guo, Kai Yu | DiffVoice: Text-to-Speech with Latent Diffusion | Accepted to ICASSP2023 | null | null | null | eess.AS cs.AI cs.HC cs.LG cs.SD | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this work, we present DiffVoice, a novel text-to-speech model based on
latent diffusion. We propose to first encode speech signals into a phoneme-rate
latent representation with a variational autoencoder enhanced by adversarial
training, and then jointly model the duration and the latent representation
with a diffusion model. Subjective evaluations on LJSpeech and LibriTTS
datasets demonstrate that our method beats the best publicly available systems
in naturalness. By adopting recent generative inverse problem solving
algorithms for diffusion models, DiffVoice achieves the state-of-the-art
performance in text-based speech editing, and zero-shot adaptation.
| [
{
"version": "v1",
"created": "Sun, 23 Apr 2023 21:05:33 GMT"
}
] | 2023-04-25T00:00:00 | [
[
"Liu",
"Zhijun",
""
],
[
"Guo",
"Yiwei",
""
],
[
"Yu",
"Kai",
""
]
] | TITLE: DiffVoice: Text-to-Speech with Latent Diffusion
ABSTRACT: In this work, we present DiffVoice, a novel text-to-speech model based on
latent diffusion. We propose to first encode speech signals into a phoneme-rate
latent representation with a variational autoencoder enhanced by adversarial
training, and then jointly model the duration and the latent representation
with a diffusion model. Subjective evaluations on LJSpeech and LibriTTS
datasets demonstrate that our method beats the best publicly available systems
in naturalness. By adopting recent generative inverse problem solving
algorithms for diffusion models, DiffVoice achieves the state-of-the-art
performance in text-based speech editing, and zero-shot adaptation.
| no_new_dataset | 0.711186 |
2304.11752 | Debasis Ganguly | Debasis Ganguly, Emine Yilmaz | Query-specific Variable Depth Pooling via Query Performance Prediction
towards Reducing Relevance Assessment Effort | To appear in SIGIR 2023 | null | null | null | cs.IR | http://creativecommons.org/licenses/by/4.0/ | Due to the massive size of test collections, a standard practice in IR
evaluation is to construct a 'pool' of candidate relevant documents comprised
of the top-k documents retrieved by a wide range of different retrieval systems
- a process called depth-k pooling. A standard practice is to set the depth (k)
to a constant value for each query constituting the benchmark set. However, in
this paper we argue that the annotation effort can be substantially reduced if
the depth of the pool is made a variable quantity for each query, the rationale
being that the number of documents relevant to the information need can widely
vary across queries. Our hypothesis is that a lower depth for the former class
of queries and a higher depth for the latter can potentially reduce the
annotation effort without a significant change in retrieval effectiveness
evaluation. We make use of standard query performance prediction (QPP)
techniques to estimate the number of potentially relevant documents for each
query, which is then used to determine the depth of the pool. Our experiments
conducted on standard test collections demonstrate that this proposed method of
employing query-specific variable depths is able to adequately reflect the
relative effectiveness of IR systems with a substantially smaller annotation
effort.
| [
{
"version": "v1",
"created": "Sun, 23 Apr 2023 21:36:49 GMT"
}
] | 2023-04-25T00:00:00 | [
[
"Ganguly",
"Debasis",
""
],
[
"Yilmaz",
"Emine",
""
]
] | TITLE: Query-specific Variable Depth Pooling via Query Performance Prediction
towards Reducing Relevance Assessment Effort
ABSTRACT: Due to the massive size of test collections, a standard practice in IR
evaluation is to construct a 'pool' of candidate relevant documents comprised
of the top-k documents retrieved by a wide range of different retrieval systems
- a process called depth-k pooling. A standard practice is to set the depth (k)
to a constant value for each query constituting the benchmark set. However, in
this paper we argue that the annotation effort can be substantially reduced if
the depth of the pool is made a variable quantity for each query, the rationale
being that the number of documents relevant to the information need can widely
vary across queries. Our hypothesis is that a lower depth for the former class
of queries and a higher depth for the latter can potentially reduce the
annotation effort without a significant change in retrieval effectiveness
evaluation. We make use of standard query performance prediction (QPP)
techniques to estimate the number of potentially relevant documents for each
query, which is then used to determine the depth of the pool. Our experiments
conducted on standard test collections demonstrate that this proposed method of
employing query-specific variable depths is able to adequately reflect the
relative effectiveness of IR systems with a substantially smaller annotation
effort.
| no_new_dataset | 0.711663 |
2304.11758 | Oleg I. Berngardt | Oleg I.Berngardt | Improving Classification Neural Networks by using Absolute activation
function (MNIST/LeNET-5 example) | 19 pages, 5 figures, 3 tables, submitted to Pattern Recognition | null | null | null | cs.LG cs.CV | http://creativecommons.org/licenses/by/4.0/ | The paper discusses the use of the Absolute activation function in
classification neural networks. An examples are shown of using this activation
function in simple and more complex problems. Using as a baseline LeNet-5
network for solving the MNIST problem, the efficiency of Absolute activation
function is shown in comparison with the use of Tanh, ReLU and SeLU
activations. It is shown that in deep networks Absolute activation does not
cause vanishing and exploding gradients, and therefore Absolute activation can
be used in both simple and deep neural networks. Due to high volatility of
training networks with Absolute activation, a special modification of ADAM
training algorithm is used, that estimates lower bound of accuracy at any test
dataset using validation dataset analysis at each training epoch, and uses this
value to stop/decrease learning rate, and re-initializes ADAM algorithm between
these steps. It is shown that solving the MNIST problem with the LeNet-like
architectures based on Absolute activation allows to significantly reduce the
number of trained parameters in the neural network with improving the
prediction accuracy.
| [
{
"version": "v1",
"created": "Sun, 23 Apr 2023 22:17:58 GMT"
}
] | 2023-04-25T00:00:00 | [
[
"Berngardt",
"Oleg I.",
""
]
] | TITLE: Improving Classification Neural Networks by using Absolute activation
function (MNIST/LeNET-5 example)
ABSTRACT: The paper discusses the use of the Absolute activation function in
classification neural networks. An examples are shown of using this activation
function in simple and more complex problems. Using as a baseline LeNet-5
network for solving the MNIST problem, the efficiency of Absolute activation
function is shown in comparison with the use of Tanh, ReLU and SeLU
activations. It is shown that in deep networks Absolute activation does not
cause vanishing and exploding gradients, and therefore Absolute activation can
be used in both simple and deep neural networks. Due to high volatility of
training networks with Absolute activation, a special modification of ADAM
training algorithm is used, that estimates lower bound of accuracy at any test
dataset using validation dataset analysis at each training epoch, and uses this
value to stop/decrease learning rate, and re-initializes ADAM algorithm between
these steps. It is shown that solving the MNIST problem with the LeNet-like
architectures based on Absolute activation allows to significantly reduce the
number of trained parameters in the neural network with improving the
prediction accuracy.
| no_new_dataset | 0.711281 |
2304.11788 | Hongchang Gao | Yihan Zhang, Wenhao Jiang, Feng Zheng, Chiu C. Tan, Xinghua Shi,
Hongchang Gao | Can Decentralized Stochastic Minimax Optimization Algorithms Converge
Linearly for Finite-Sum Nonconvex-Nonconcave Problems? | Under submission | null | null | null | cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Decentralized minimax optimization has been actively studied in the past few
years due to its application in a wide range of machine learning models.
However, the current theoretical understanding of its convergence rate is far
from satisfactory since existing works only focus on the
nonconvex-strongly-concave problem. This motivates us to study decentralized
minimax optimization algorithms for the nonconvex-nonconcave problem. To this
end, we develop two novel decentralized stochastic variance-reduced gradient
descent ascent algorithms for the finite-sum nonconvex-nonconcave problem that
satisfies the Polyak-{\L}ojasiewicz (PL) condition. In particular, our
theoretical analyses demonstrate how to conduct local updates and perform
communication to achieve the linear convergence rate. To the best of our
knowledge, this is the first work achieving linear convergence rates for
decentralized nonconvex-nonconcave problems. Finally, we verify the performance
of our algorithms on both synthetic and real-world datasets. The experimental
results confirm the efficacy of our algorithms.
| [
{
"version": "v1",
"created": "Mon, 24 Apr 2023 02:19:39 GMT"
}
] | 2023-04-25T00:00:00 | [
[
"Zhang",
"Yihan",
""
],
[
"Jiang",
"Wenhao",
""
],
[
"Zheng",
"Feng",
""
],
[
"Tan",
"Chiu C.",
""
],
[
"Shi",
"Xinghua",
""
],
[
"Gao",
"Hongchang",
""
]
] | TITLE: Can Decentralized Stochastic Minimax Optimization Algorithms Converge
Linearly for Finite-Sum Nonconvex-Nonconcave Problems?
ABSTRACT: Decentralized minimax optimization has been actively studied in the past few
years due to its application in a wide range of machine learning models.
However, the current theoretical understanding of its convergence rate is far
from satisfactory since existing works only focus on the
nonconvex-strongly-concave problem. This motivates us to study decentralized
minimax optimization algorithms for the nonconvex-nonconcave problem. To this
end, we develop two novel decentralized stochastic variance-reduced gradient
descent ascent algorithms for the finite-sum nonconvex-nonconcave problem that
satisfies the Polyak-{\L}ojasiewicz (PL) condition. In particular, our
theoretical analyses demonstrate how to conduct local updates and perform
communication to achieve the linear convergence rate. To the best of our
knowledge, this is the first work achieving linear convergence rates for
decentralized nonconvex-nonconcave problems. Finally, we verify the performance
of our algorithms on both synthetic and real-world datasets. The experimental
results confirm the efficacy of our algorithms.
| no_new_dataset | 0.708364 |
2304.11790 | Khoi Nguyen-Duy Minh | Khoi Minh Nguyen-Duy, Quang Pham, Binh T. Nguyen | Adaptive-saturated RNN: Remember more with less instability | 8 pages, 2 figures, 5 tables, ICLR 2023 Tiny Paper Track | ICLR 2023 Tiny Paper Track | null | null | cs.LG | http://creativecommons.org/licenses/by/4.0/ | Orthogonal parameterization is a compelling solution to the vanishing
gradient problem (VGP) in recurrent neural networks (RNNs). With orthogonal
parameters and non-saturated activation functions, gradients in such models are
constrained to unit norms. On the other hand, although the traditional vanilla
RNNs are seen to have higher memory capacity, they suffer from the VGP and
perform badly in many applications. This work proposes Adaptive-Saturated RNNs
(asRNN), a variant that dynamically adjusts its saturation level between the
two mentioned approaches. Consequently, asRNN enjoys both the capacity of a
vanilla RNN and the training stability of orthogonal RNNs. Our experiments show
encouraging results of asRNN on challenging sequence learning benchmarks
compared to several strong competitors. The research code is accessible at
https://github.com/ndminhkhoi46/asRNN/.
| [
{
"version": "v1",
"created": "Mon, 24 Apr 2023 02:28:03 GMT"
}
] | 2023-04-25T00:00:00 | [
[
"Nguyen-Duy",
"Khoi Minh",
""
],
[
"Pham",
"Quang",
""
],
[
"Nguyen",
"Binh T.",
""
]
] | TITLE: Adaptive-saturated RNN: Remember more with less instability
ABSTRACT: Orthogonal parameterization is a compelling solution to the vanishing
gradient problem (VGP) in recurrent neural networks (RNNs). With orthogonal
parameters and non-saturated activation functions, gradients in such models are
constrained to unit norms. On the other hand, although the traditional vanilla
RNNs are seen to have higher memory capacity, they suffer from the VGP and
perform badly in many applications. This work proposes Adaptive-Saturated RNNs
(asRNN), a variant that dynamically adjusts its saturation level between the
two mentioned approaches. Consequently, asRNN enjoys both the capacity of a
vanilla RNN and the training stability of orthogonal RNNs. Our experiments show
encouraging results of asRNN on challenging sequence learning benchmarks
compared to several strong competitors. The research code is accessible at
https://github.com/ndminhkhoi46/asRNN/.
| no_new_dataset | 0.710377 |
2304.11805 | Xuexue Li | Xuexue Li, Wenhui Diao, Yongqiang Mao, Peng Gao, Xiuhua Mao, Xinming
Li and Xian Sun | OGMN: Occlusion-guided Multi-task Network for Object Detection in UAV
Images | 20 pages, 15 figures | null | 10.1016/j.isprsjprs.2023.04.009 | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Occlusion between objects is one of the overlooked challenges for object
detection in UAV images. Due to the variable altitude and angle of UAVs,
occlusion in UAV images happens more frequently than that in natural scenes.
Compared to occlusion in natural scene images, occlusion in UAV images happens
with feature confusion problem and local aggregation characteristic. And we
found that extracting or localizing occlusion between objects is beneficial for
the detector to address this challenge. According to this finding, the
occlusion localization task is introduced, which together with the object
detection task constitutes our occlusion-guided multi-task network (OGMN). The
OGMN contains the localization of occlusion and two occlusion-guided multi-task
interactions. In detail, an occlusion estimation module (OEM) is proposed to
precisely localize occlusion. Then the OGMN utilizes the occlusion localization
results to implement occlusion-guided detection with two multi-task
interactions. One interaction for the guide is between two task decoders to
address the feature confusion problem, and an occlusion decoupling head (ODH)
is proposed to replace the general detection head. Another interaction for
guide is designed in the detection process according to local aggregation
characteristic, and a two-phase progressive refinement process (TPP) is
proposed to optimize the detection process. Extensive experiments demonstrate
the effectiveness of our OGMN on the Visdrone and UAVDT datasets. In
particular, our OGMN achieves 35.0% mAP on the Visdrone dataset and outperforms
the baseline by 5.3%. And our OGMN provides a new insight for accurate
occlusion localization and achieves competitive detection performance.
| [
{
"version": "v1",
"created": "Mon, 24 Apr 2023 03:30:00 GMT"
}
] | 2023-04-25T00:00:00 | [
[
"Li",
"Xuexue",
""
],
[
"Diao",
"Wenhui",
""
],
[
"Mao",
"Yongqiang",
""
],
[
"Gao",
"Peng",
""
],
[
"Mao",
"Xiuhua",
""
],
[
"Li",
"Xinming",
""
],
[
"Sun",
"Xian",
""
]
] | TITLE: OGMN: Occlusion-guided Multi-task Network for Object Detection in UAV
Images
ABSTRACT: Occlusion between objects is one of the overlooked challenges for object
detection in UAV images. Due to the variable altitude and angle of UAVs,
occlusion in UAV images happens more frequently than that in natural scenes.
Compared to occlusion in natural scene images, occlusion in UAV images happens
with feature confusion problem and local aggregation characteristic. And we
found that extracting or localizing occlusion between objects is beneficial for
the detector to address this challenge. According to this finding, the
occlusion localization task is introduced, which together with the object
detection task constitutes our occlusion-guided multi-task network (OGMN). The
OGMN contains the localization of occlusion and two occlusion-guided multi-task
interactions. In detail, an occlusion estimation module (OEM) is proposed to
precisely localize occlusion. Then the OGMN utilizes the occlusion localization
results to implement occlusion-guided detection with two multi-task
interactions. One interaction for the guide is between two task decoders to
address the feature confusion problem, and an occlusion decoupling head (ODH)
is proposed to replace the general detection head. Another interaction for
guide is designed in the detection process according to local aggregation
characteristic, and a two-phase progressive refinement process (TPP) is
proposed to optimize the detection process. Extensive experiments demonstrate
the effectiveness of our OGMN on the Visdrone and UAVDT datasets. In
particular, our OGMN achieves 35.0% mAP on the Visdrone dataset and outperforms
the baseline by 5.3%. And our OGMN provides a new insight for accurate
occlusion localization and achieves competitive detection performance.
| no_new_dataset | 0.711462 |
2304.11810 | Shu Wei | Shu Wei and Nuo Xu | PARAGRAPH2GRAPH: A GNN-based framework for layout paragraph analysis | null | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | Document layout analysis has a wide range of requirements across various
domains, languages, and business scenarios. However, most current
state-of-the-art algorithms are language-dependent, with architectures that
rely on transformer encoders or language-specific text encoders, such as BERT,
for feature extraction. These approaches are limited in their ability to handle
very long documents due to input sequence length constraints and are closely
tied to language-specific tokenizers. Additionally, training a cross-language
text encoder can be challenging due to the lack of labeled multilingual
document datasets that consider privacy. Furthermore, some layout tasks require
a clean separation between different layout components without overlap, which
can be difficult for image segmentation-based algorithms to achieve. In this
paper, we present Paragraph2Graph, a language-independent graph neural network
(GNN)-based model that achieves competitive results on common document layout
datasets while being adaptable to business scenarios with strict separation.
With only 19.95 million parameters, our model is suitable for industrial
applications, particularly in multi-language scenarios.
| [
{
"version": "v1",
"created": "Mon, 24 Apr 2023 03:54:48 GMT"
}
] | 2023-04-25T00:00:00 | [
[
"Wei",
"Shu",
""
],
[
"Xu",
"Nuo",
""
]
] | TITLE: PARAGRAPH2GRAPH: A GNN-based framework for layout paragraph analysis
ABSTRACT: Document layout analysis has a wide range of requirements across various
domains, languages, and business scenarios. However, most current
state-of-the-art algorithms are language-dependent, with architectures that
rely on transformer encoders or language-specific text encoders, such as BERT,
for feature extraction. These approaches are limited in their ability to handle
very long documents due to input sequence length constraints and are closely
tied to language-specific tokenizers. Additionally, training a cross-language
text encoder can be challenging due to the lack of labeled multilingual
document datasets that consider privacy. Furthermore, some layout tasks require
a clean separation between different layout components without overlap, which
can be difficult for image segmentation-based algorithms to achieve. In this
paper, we present Paragraph2Graph, a language-independent graph neural network
(GNN)-based model that achieves competitive results on common document layout
datasets while being adaptable to business scenarios with strict separation.
With only 19.95 million parameters, our model is suitable for industrial
applications, particularly in multi-language scenarios.
| no_new_dataset | 0.708187 |
2304.11812 | Guangzhe Hou | Guangzhe Hou, Guihe Qin, Minghui Sun, Yanhua Liang, Jie Yan, Zhonghan
Zhang | NoiseTrans: Point Cloud Denoising with Transformers | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Point clouds obtained from capture devices or 3D reconstruction techniques
are often noisy and interfere with downstream tasks. The paper aims to recover
the underlying surface of noisy point clouds. We design a novel model,
NoiseTrans, which uses transformer encoder architecture for point cloud
denoising. Specifically, we obtain structural similarity of point-based point
clouds with the assistance of the transformer's core self-attention mechanism.
By expressing the noisy point cloud as a set of unordered vectors, we convert
point clouds into point embeddings and employ Transformer to generate clean
point clouds. To make the Transformer preserve details when sensing the point
cloud, we design the Local Point Attention to prevent the point cloud from
being over-smooth. In addition, we also propose sparse encoding, which enables
the Transformer to better perceive the structural relationships of the point
cloud and improve the denoising performance. Experiments show that our model
outperforms state-of-the-art methods in various datasets and noise
environments.
| [
{
"version": "v1",
"created": "Mon, 24 Apr 2023 04:01:23 GMT"
}
] | 2023-04-25T00:00:00 | [
[
"Hou",
"Guangzhe",
""
],
[
"Qin",
"Guihe",
""
],
[
"Sun",
"Minghui",
""
],
[
"Liang",
"Yanhua",
""
],
[
"Yan",
"Jie",
""
],
[
"Zhang",
"Zhonghan",
""
]
] | TITLE: NoiseTrans: Point Cloud Denoising with Transformers
ABSTRACT: Point clouds obtained from capture devices or 3D reconstruction techniques
are often noisy and interfere with downstream tasks. The paper aims to recover
the underlying surface of noisy point clouds. We design a novel model,
NoiseTrans, which uses transformer encoder architecture for point cloud
denoising. Specifically, we obtain structural similarity of point-based point
clouds with the assistance of the transformer's core self-attention mechanism.
By expressing the noisy point cloud as a set of unordered vectors, we convert
point clouds into point embeddings and employ Transformer to generate clean
point clouds. To make the Transformer preserve details when sensing the point
cloud, we design the Local Point Attention to prevent the point cloud from
being over-smooth. In addition, we also propose sparse encoding, which enables
the Transformer to better perceive the structural relationships of the point
cloud and improve the denoising performance. Experiments show that our model
outperforms state-of-the-art methods in various datasets and noise
environments.
| no_new_dataset | 0.70825 |
2304.11846 | Yun He | Yun He and Danhang Tang and Yinda Zhang and Xiangyang Xue and Yanwei
Fu | Grad-PU: Arbitrary-Scale Point Cloud Upsampling via Gradient Descent
with Learned Distance Functions | Accepted by CVPR 2023. Code is avaliable at
https://github.com/yunhe20/Grad-PU | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Most existing point cloud upsampling methods have roughly three steps:
feature extraction, feature expansion and 3D coordinate prediction.
However,they usually suffer from two critical issues: (1)fixed upsampling rate
after one-time training, since the feature expansion unit is customized for
each upsampling rate; (2)outliers or shrinkage artifact caused by the
difficulty of precisely predicting 3D coordinates or residuals of upsampled
points. To adress them, we propose a new framework for accurate point cloud
upsampling that supports arbitrary upsampling rates. Our method first
interpolates the low-res point cloud according to a given upsampling rate. And
then refine the positions of the interpolated points with an iterative
optimization process, guided by a trained model estimating the difference
between the current point cloud and the high-res target. Extensive quantitative
and qualitative results on benchmarks and downstream tasks demonstrate that our
method achieves the state-of-the-art accuracy and efficiency.
| [
{
"version": "v1",
"created": "Mon, 24 Apr 2023 06:36:35 GMT"
}
] | 2023-04-25T00:00:00 | [
[
"He",
"Yun",
""
],
[
"Tang",
"Danhang",
""
],
[
"Zhang",
"Yinda",
""
],
[
"Xue",
"Xiangyang",
""
],
[
"Fu",
"Yanwei",
""
]
] | TITLE: Grad-PU: Arbitrary-Scale Point Cloud Upsampling via Gradient Descent
with Learned Distance Functions
ABSTRACT: Most existing point cloud upsampling methods have roughly three steps:
feature extraction, feature expansion and 3D coordinate prediction.
However,they usually suffer from two critical issues: (1)fixed upsampling rate
after one-time training, since the feature expansion unit is customized for
each upsampling rate; (2)outliers or shrinkage artifact caused by the
difficulty of precisely predicting 3D coordinates or residuals of upsampled
points. To adress them, we propose a new framework for accurate point cloud
upsampling that supports arbitrary upsampling rates. Our method first
interpolates the low-res point cloud according to a given upsampling rate. And
then refine the positions of the interpolated points with an iterative
optimization process, guided by a trained model estimating the difference
between the current point cloud and the high-res target. Extensive quantitative
and qualitative results on benchmarks and downstream tasks demonstrate that our
method achieves the state-of-the-art accuracy and efficiency.
| no_new_dataset | 0.710258 |
2304.11855 | Haoyu Wang | Haoyu Wang, Guansong Pang, Peng Wang, Lei Zhang, Wei Wei, Yanning
Zhang | Glocal Energy-based Learning for Few-Shot Open-Set Recognition | Accepted at CVPR 2023 | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Few-shot open-set recognition (FSOR) is a challenging task of great practical
value. It aims to categorize a sample to one of the pre-defined, closed-set
classes illustrated by few examples while being able to reject the sample from
unknown classes. In this work, we approach the FSOR task by proposing a novel
energy-based hybrid model. The model is composed of two branches, where a
classification branch learns a metric to classify a sample to one of closed-set
classes and the energy branch explicitly estimates the open-set probability. To
achieve holistic detection of open-set samples, our model leverages both
class-wise and pixel-wise features to learn a glocal energy-based score, in
which a global energy score is learned using the class-wise features, while a
local energy score is learned using the pixel-wise features. The model is
enforced to assign large energy scores to samples that are deviated from the
few-shot examples in either the class-wise features or the pixel-wise features,
and to assign small energy scores otherwise. Experiments on three standard FSOR
datasets show the superior performance of our model.
| [
{
"version": "v1",
"created": "Mon, 24 Apr 2023 07:06:50 GMT"
}
] | 2023-04-25T00:00:00 | [
[
"Wang",
"Haoyu",
""
],
[
"Pang",
"Guansong",
""
],
[
"Wang",
"Peng",
""
],
[
"Zhang",
"Lei",
""
],
[
"Wei",
"Wei",
""
],
[
"Zhang",
"Yanning",
""
]
] | TITLE: Glocal Energy-based Learning for Few-Shot Open-Set Recognition
ABSTRACT: Few-shot open-set recognition (FSOR) is a challenging task of great practical
value. It aims to categorize a sample to one of the pre-defined, closed-set
classes illustrated by few examples while being able to reject the sample from
unknown classes. In this work, we approach the FSOR task by proposing a novel
energy-based hybrid model. The model is composed of two branches, where a
classification branch learns a metric to classify a sample to one of closed-set
classes and the energy branch explicitly estimates the open-set probability. To
achieve holistic detection of open-set samples, our model leverages both
class-wise and pixel-wise features to learn a glocal energy-based score, in
which a global energy score is learned using the class-wise features, while a
local energy score is learned using the pixel-wise features. The model is
enforced to assign large energy scores to samples that are deviated from the
few-shot examples in either the class-wise features or the pixel-wise features,
and to assign small energy scores otherwise. Experiments on three standard FSOR
datasets show the superior performance of our model.
| no_new_dataset | 0.713251 |
2304.11947 | Zefeng Chen | Zefeng Chen, Wensheng Gan, Gengsen Huang, Yan Li, and Zhenlian Qi | Towards Top-$K$ Non-Overlapping Sequential Patterns | Preprint. 5 figures, 5 tables | null | null | null | cs.DB | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Sequential pattern mining (SPM) has excellent prospects and application
spaces and has been widely used in different fields. The non-overlapping SPM,
as one of the data mining techniques, has been used to discover patterns that
have requirements for gap constraints in some specific mining tasks, such as
bio-data mining. And for the non-overlapping sequential patterns with gap
constraints, the Nettree structure has been proposed to efficiently compute the
support of the patterns. For pattern mining, users usually need to consider the
threshold of minimum support (\textit{minsup}). This is especially difficult in
the case of large databases. Although some existing algorithms can mine the
top-$k$ patterns, they are approximate algorithms with fixed lengths. In this
paper, a precise algorithm for mining \underline{T}op-$k$
\underline{N}on-\underline{O}verlapping \underline{S}equential
\underline{P}atterns (TNOSP) is proposed. The top-$k$ solution of SPM is an
effective way to discover the most frequent non-overlapping sequential patterns
without having to set the \textit{minsup}. As a novel pattern mining algorithm,
TNOSP can precisely search the top-$k$ patterns of non-overlapping sequences
with different gap constraints. We further propose a pruning strategy named
\underline{Q}ueue \underline{M}eta \underline{S}et \underline{P}runing (QMSP)
to improve TNOSP's performance. TNOSP can reduce redundancy in non-overlapping
sequential mining and has better performance in mining precise non-overlapping
sequential patterns. The experimental results and comparisons on several
datasets have shown that TNOSP outperformed the existing algorithms in terms of
precision, efficiency, and scalability.
| [
{
"version": "v1",
"created": "Mon, 24 Apr 2023 09:32:09 GMT"
}
] | 2023-04-25T00:00:00 | [
[
"Chen",
"Zefeng",
""
],
[
"Gan",
"Wensheng",
""
],
[
"Huang",
"Gengsen",
""
],
[
"Li",
"Yan",
""
],
[
"Qi",
"Zhenlian",
""
]
] | TITLE: Towards Top-$K$ Non-Overlapping Sequential Patterns
ABSTRACT: Sequential pattern mining (SPM) has excellent prospects and application
spaces and has been widely used in different fields. The non-overlapping SPM,
as one of the data mining techniques, has been used to discover patterns that
have requirements for gap constraints in some specific mining tasks, such as
bio-data mining. And for the non-overlapping sequential patterns with gap
constraints, the Nettree structure has been proposed to efficiently compute the
support of the patterns. For pattern mining, users usually need to consider the
threshold of minimum support (\textit{minsup}). This is especially difficult in
the case of large databases. Although some existing algorithms can mine the
top-$k$ patterns, they are approximate algorithms with fixed lengths. In this
paper, a precise algorithm for mining \underline{T}op-$k$
\underline{N}on-\underline{O}verlapping \underline{S}equential
\underline{P}atterns (TNOSP) is proposed. The top-$k$ solution of SPM is an
effective way to discover the most frequent non-overlapping sequential patterns
without having to set the \textit{minsup}. As a novel pattern mining algorithm,
TNOSP can precisely search the top-$k$ patterns of non-overlapping sequences
with different gap constraints. We further propose a pruning strategy named
\underline{Q}ueue \underline{M}eta \underline{S}et \underline{P}runing (QMSP)
to improve TNOSP's performance. TNOSP can reduce redundancy in non-overlapping
sequential mining and has better performance in mining precise non-overlapping
sequential patterns. The experimental results and comparisons on several
datasets have shown that TNOSP outperformed the existing algorithms in terms of
precision, efficiency, and scalability.
| no_new_dataset | 0.711055 |
2304.11955 | Yaxin Shi | Yaxin Shi, Xiaowei Zhou, Ping Liu, Ivor W. Tsang | UTSGAN: Unseen Transition Suss GAN for Transition-Aware Image-to-image
Translation | 17 pages, 17 figures | null | null | null | cs.CV cs.LG | http://creativecommons.org/licenses/by-nc-nd/4.0/ | In the field of Image-to-Image (I2I) translation, ensuring consistency
between input images and their translated results is a key requirement for
producing high-quality and desirable outputs. Previous I2I methods have relied
on result consistency, which enforces consistency between the translated
results and the ground truth output, to achieve this goal. However, result
consistency is limited in its ability to handle complex and unseen attribute
changes in translation tasks. To address this issue, we introduce a
transition-aware approach to I2I translation, where the data translation
mapping is explicitly parameterized with a transition variable, allowing for
the modelling of unobserved translations triggered by unseen transitions.
Furthermore, we propose the use of transition consistency, defined on the
transition variable, to enable regularization of consistency on unobserved
translations, which is omitted in previous works. Based on these insights, we
present Unseen Transition Suss GAN (UTSGAN), a generative framework that
constructs a manifold for the transition with a stochastic transition encoder
and coherently regularizes and generalizes result consistency and transition
consistency on both training and unobserved translations with tailor-designed
constraints. Extensive experiments on four different I2I tasks performed on
five different datasets demonstrate the efficacy of our proposed UTSGAN in
performing consistent translations.
| [
{
"version": "v1",
"created": "Mon, 24 Apr 2023 09:47:34 GMT"
}
] | 2023-04-25T00:00:00 | [
[
"Shi",
"Yaxin",
""
],
[
"Zhou",
"Xiaowei",
""
],
[
"Liu",
"Ping",
""
],
[
"Tsang",
"Ivor W.",
""
]
] | TITLE: UTSGAN: Unseen Transition Suss GAN for Transition-Aware Image-to-image
Translation
ABSTRACT: In the field of Image-to-Image (I2I) translation, ensuring consistency
between input images and their translated results is a key requirement for
producing high-quality and desirable outputs. Previous I2I methods have relied
on result consistency, which enforces consistency between the translated
results and the ground truth output, to achieve this goal. However, result
consistency is limited in its ability to handle complex and unseen attribute
changes in translation tasks. To address this issue, we introduce a
transition-aware approach to I2I translation, where the data translation
mapping is explicitly parameterized with a transition variable, allowing for
the modelling of unobserved translations triggered by unseen transitions.
Furthermore, we propose the use of transition consistency, defined on the
transition variable, to enable regularization of consistency on unobserved
translations, which is omitted in previous works. Based on these insights, we
present Unseen Transition Suss GAN (UTSGAN), a generative framework that
constructs a manifold for the transition with a stochastic transition encoder
and coherently regularizes and generalizes result consistency and transition
consistency on both training and unobserved translations with tailor-designed
constraints. Extensive experiments on four different I2I tasks performed on
five different datasets demonstrate the efficacy of our proposed UTSGAN in
performing consistent translations.
| no_new_dataset | 0.709648 |
2304.11969 | Ziqi Xu | Ziqi Xu, Debo Cheng, Jiuyong Li, Jixue Liu, Lin Liu, Kui Yu | Causal Effect Estimation with Variational AutoEncoder and the Front Door
Criterion | null | null | null | null | cs.LG stat.ME | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | An essential problem in causal inference is estimating causal effects from
observational data. The problem becomes more challenging with the presence of
unobserved confounders. When there are unobserved confounders, the commonly
used back-door adjustment is not applicable. Although the instrumental variable
(IV) methods can deal with unobserved confounders, they all assume that the
treatment directly affects the outcome, and there is no mediator between the
treatment and the outcome. This paper aims to use the front-door criterion to
address the challenging problem with the presence of unobserved confounders and
mediators. In practice, it is often difficult to identify the set of variables
used for front-door adjustment from data. By leveraging the ability of deep
generative models in representation learning, we propose FDVAE to learn the
representation of a Front-Door adjustment set with a Variational AutoEncoder,
instead of trying to search for a set of variables for front-door adjustment.
Extensive experiments on synthetic datasets validate the effectiveness of FDVAE
and its superiority over existing methods. The experiments also show that the
performance of FDVAE is not sensitive to the causal strength of unobserved
confounders and is feasible in the case of dimensionality mismatch between
learned representations and the ground truth. We further apply the method to
three real-world datasets to demonstrate its potential applications.
| [
{
"version": "v1",
"created": "Mon, 24 Apr 2023 10:04:28 GMT"
}
] | 2023-04-25T00:00:00 | [
[
"Xu",
"Ziqi",
""
],
[
"Cheng",
"Debo",
""
],
[
"Li",
"Jiuyong",
""
],
[
"Liu",
"Jixue",
""
],
[
"Liu",
"Lin",
""
],
[
"Yu",
"Kui",
""
]
] | TITLE: Causal Effect Estimation with Variational AutoEncoder and the Front Door
Criterion
ABSTRACT: An essential problem in causal inference is estimating causal effects from
observational data. The problem becomes more challenging with the presence of
unobserved confounders. When there are unobserved confounders, the commonly
used back-door adjustment is not applicable. Although the instrumental variable
(IV) methods can deal with unobserved confounders, they all assume that the
treatment directly affects the outcome, and there is no mediator between the
treatment and the outcome. This paper aims to use the front-door criterion to
address the challenging problem with the presence of unobserved confounders and
mediators. In practice, it is often difficult to identify the set of variables
used for front-door adjustment from data. By leveraging the ability of deep
generative models in representation learning, we propose FDVAE to learn the
representation of a Front-Door adjustment set with a Variational AutoEncoder,
instead of trying to search for a set of variables for front-door adjustment.
Extensive experiments on synthetic datasets validate the effectiveness of FDVAE
and its superiority over existing methods. The experiments also show that the
performance of FDVAE is not sensitive to the causal strength of unobserved
confounders and is feasible in the case of dimensionality mismatch between
learned representations and the ground truth. We further apply the method to
three real-world datasets to demonstrate its potential applications.
| no_new_dataset | 0.709435 |
2304.11970 | Zerui Chen | Zerui Chen, Shizhe Chen, Cordelia Schmid, Ivan Laptev | gSDF: Geometry-Driven Signed Distance Functions for 3D Hand-Object
Reconstruction | Accepted by CVPR 2023. Project Page:
https://zerchen.github.io/projects/gsdf.html | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Signed distance functions (SDFs) is an attractive framework that has recently
shown promising results for 3D shape reconstruction from images. SDFs
seamlessly generalize to different shape resolutions and topologies but lack
explicit modelling of the underlying 3D geometry. In this work, we exploit the
hand structure and use it as guidance for SDF-based shape reconstruction. In
particular, we address reconstruction of hands and manipulated objects from
monocular RGB images. To this end, we estimate poses of hands and objects and
use them to guide 3D reconstruction. More specifically, we predict kinematic
chains of pose transformations and align SDFs with highly-articulated hand
poses. We improve the visual features of 3D points with geometry alignment and
further leverage temporal information to enhance the robustness to occlusion
and motion blurs. We conduct extensive experiments on the challenging ObMan and
DexYCB benchmarks and demonstrate significant improvements of the proposed
method over the state of the art.
| [
{
"version": "v1",
"created": "Mon, 24 Apr 2023 10:05:48 GMT"
}
] | 2023-04-25T00:00:00 | [
[
"Chen",
"Zerui",
""
],
[
"Chen",
"Shizhe",
""
],
[
"Schmid",
"Cordelia",
""
],
[
"Laptev",
"Ivan",
""
]
] | TITLE: gSDF: Geometry-Driven Signed Distance Functions for 3D Hand-Object
Reconstruction
ABSTRACT: Signed distance functions (SDFs) is an attractive framework that has recently
shown promising results for 3D shape reconstruction from images. SDFs
seamlessly generalize to different shape resolutions and topologies but lack
explicit modelling of the underlying 3D geometry. In this work, we exploit the
hand structure and use it as guidance for SDF-based shape reconstruction. In
particular, we address reconstruction of hands and manipulated objects from
monocular RGB images. To this end, we estimate poses of hands and objects and
use them to guide 3D reconstruction. More specifically, we predict kinematic
chains of pose transformations and align SDFs with highly-articulated hand
poses. We improve the visual features of 3D points with geometry alignment and
further leverage temporal information to enhance the robustness to occlusion
and motion blurs. We conduct extensive experiments on the challenging ObMan and
DexYCB benchmarks and demonstrate significant improvements of the proposed
method over the state of the art.
| no_new_dataset | 0.711243 |
2304.11975 | Yin-Dong Zheng | Yin-Dong Zheng, Guo Chen, Minglei Yuan, Tong Lu | MRSN: Multi-Relation Support Network for Video Action Detection | 6 pages | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | Action detection is a challenging video understanding task, requiring
modeling spatio-temporal and interaction relations. Current methods usually
model actor-actor and actor-context relations separately, ignoring their
complementarity and mutual support. To solve this problem, we propose a novel
network called Multi-Relation Support Network (MRSN). In MRSN, Actor-Context
Relation Encoder (ACRE) and Actor-Actor Relation Encoder (AARE) model the
actor-context and actor-actor relation separately. Then Relation Support
Encoder (RSE) computes the supports between the two relations and performs
relation-level interactions. Finally, Relation Consensus Module (RCM) enhances
two relations with the long-term relations from the Long-term Relation Bank
(LRB) and yields a consensus. Our experiments demonstrate that modeling
relations separately and performing relation-level interactions can achieve and
outperformer state-of-the-art results on two challenging video datasets: AVA
and UCF101-24.
| [
{
"version": "v1",
"created": "Mon, 24 Apr 2023 10:15:31 GMT"
}
] | 2023-04-25T00:00:00 | [
[
"Zheng",
"Yin-Dong",
""
],
[
"Chen",
"Guo",
""
],
[
"Yuan",
"Minglei",
""
],
[
"Lu",
"Tong",
""
]
] | TITLE: MRSN: Multi-Relation Support Network for Video Action Detection
ABSTRACT: Action detection is a challenging video understanding task, requiring
modeling spatio-temporal and interaction relations. Current methods usually
model actor-actor and actor-context relations separately, ignoring their
complementarity and mutual support. To solve this problem, we propose a novel
network called Multi-Relation Support Network (MRSN). In MRSN, Actor-Context
Relation Encoder (ACRE) and Actor-Actor Relation Encoder (AARE) model the
actor-context and actor-actor relation separately. Then Relation Support
Encoder (RSE) computes the supports between the two relations and performs
relation-level interactions. Finally, Relation Consensus Module (RCM) enhances
two relations with the long-term relations from the Long-term Relation Bank
(LRB) and yields a consensus. Our experiments demonstrate that modeling
relations separately and performing relation-level interactions can achieve and
outperformer state-of-the-art results on two challenging video datasets: AVA
and UCF101-24.
| no_new_dataset | 0.709189 |
2304.11979 | Yan Zhou | Yan Zhou, Jie Guo, Hao Sun, Bin Song, and Fei Richard Yu | Attention-guided Multi-step Fusion: A Hierarchical Fusion Network for
Multimodal Recommendation | null | null | null | null | cs.IR cs.MM | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The main idea of multimodal recommendation is the rational utilization of the
item's multimodal information to improve the recommendation performance.
Previous works directly integrate item multimodal features with item ID
embeddings, ignoring the inherent semantic relations contained in the
multimodal features. In this paper, we propose a novel and effective
aTtention-guided Multi-step FUsion Network for multimodal recommendation, named
TMFUN. Specifically, our model first constructs modality feature graph and item
feature graph to model the latent item-item semantic structures. Then, we use
the attention module to identify inherent connections between user-item
interaction data and multimodal data, evaluate the impact of multimodal data on
different interactions, and achieve early-step fusion of item features.
Furthermore, our model optimizes item representation through the
attention-guided multi-step fusion strategy and contrastive learning to improve
recommendation performance. The extensive experiments on three real-world
datasets show that our model has superior performance compared to the
state-of-the-art models.
| [
{
"version": "v1",
"created": "Mon, 24 Apr 2023 10:25:34 GMT"
}
] | 2023-04-25T00:00:00 | [
[
"Zhou",
"Yan",
""
],
[
"Guo",
"Jie",
""
],
[
"Sun",
"Hao",
""
],
[
"Song",
"Bin",
""
],
[
"Yu",
"Fei Richard",
""
]
] | TITLE: Attention-guided Multi-step Fusion: A Hierarchical Fusion Network for
Multimodal Recommendation
ABSTRACT: The main idea of multimodal recommendation is the rational utilization of the
item's multimodal information to improve the recommendation performance.
Previous works directly integrate item multimodal features with item ID
embeddings, ignoring the inherent semantic relations contained in the
multimodal features. In this paper, we propose a novel and effective
aTtention-guided Multi-step FUsion Network for multimodal recommendation, named
TMFUN. Specifically, our model first constructs modality feature graph and item
feature graph to model the latent item-item semantic structures. Then, we use
the attention module to identify inherent connections between user-item
interaction data and multimodal data, evaluate the impact of multimodal data on
different interactions, and achieve early-step fusion of item features.
Furthermore, our model optimizes item representation through the
attention-guided multi-step fusion strategy and contrastive learning to improve
recommendation performance. The extensive experiments on three real-world
datasets show that our model has superior performance compared to the
state-of-the-art models.
| no_new_dataset | 0.709988 |
2304.11985 | Mohan Li | Mohan Li, Rama Doddipatla and Catalin Zorila | Self-regularised Minimum Latency Training for Streaming
Transformer-based Speech Recognition | 5 pages, 2 figures, accepted at Interspeech2022 | null | null | null | eess.AS cs.SD | http://creativecommons.org/licenses/by-nc-nd/4.0/ | This paper proposes a self-regularised minimum latency training (SR-MLT)
method for streaming Transformer-based automatic speech recognition (ASR)
systems. In previous works, latency was optimised by truncating the online
attention weights based on the hard alignments obtained from conventional ASR
models, without taking into account the potential loss of ASR accuracy. On the
contrary, here we present a strategy to obtain the alignments as a part of the
model training without external supervision. The alignments produced by the
proposed method are dynamically regularised on the training data, such that the
latency reduction does not result in the loss of ASR accuracy. SR-MLT is
applied as a fine-tuning step on the pre-trained Transformer models that are
based on either monotonic chunkwise attention (MoChA) or cumulative attention
(CA) algorithms for online decoding. ASR experiments on the AIShell-1 and
Librispeech datasets show that when applied on a decent pre-trained MoChA or CA
baseline model, SR-MLT can effectively reduce the latency with the relative
gains ranging from 11.8% to 39.5%. Furthermore, we also demonstrate that under
certain accuracy levels, the models trained with SR-MLT can achieve lower
latency when compared to those supervised using external hard alignments.
| [
{
"version": "v1",
"created": "Mon, 24 Apr 2023 10:38:09 GMT"
}
] | 2023-04-25T00:00:00 | [
[
"Li",
"Mohan",
""
],
[
"Doddipatla",
"Rama",
""
],
[
"Zorila",
"Catalin",
""
]
] | TITLE: Self-regularised Minimum Latency Training for Streaming
Transformer-based Speech Recognition
ABSTRACT: This paper proposes a self-regularised minimum latency training (SR-MLT)
method for streaming Transformer-based automatic speech recognition (ASR)
systems. In previous works, latency was optimised by truncating the online
attention weights based on the hard alignments obtained from conventional ASR
models, without taking into account the potential loss of ASR accuracy. On the
contrary, here we present a strategy to obtain the alignments as a part of the
model training without external supervision. The alignments produced by the
proposed method are dynamically regularised on the training data, such that the
latency reduction does not result in the loss of ASR accuracy. SR-MLT is
applied as a fine-tuning step on the pre-trained Transformer models that are
based on either monotonic chunkwise attention (MoChA) or cumulative attention
(CA) algorithms for online decoding. ASR experiments on the AIShell-1 and
Librispeech datasets show that when applied on a decent pre-trained MoChA or CA
baseline model, SR-MLT can effectively reduce the latency with the relative
gains ranging from 11.8% to 39.5%. Furthermore, we also demonstrate that under
certain accuracy levels, the models trained with SR-MLT can achieve lower
latency when compared to those supervised using external hard alignments.
| no_new_dataset | 0.710622 |
2304.11989 | Yinchuan Li | Yinchuan Li, Zhigang Li, Wenqian Li, Yunfeng Shao, Yan Zheng and
Jianye Hao | Generative Flow Networks for Precise Reward-Oriented Active Learning on
Graphs | null | null | null | null | cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Many score-based active learning methods have been successfully applied to
graph-structured data, aiming to reduce the number of labels and achieve better
performance of graph neural networks based on predefined score functions.
However, these algorithms struggle to learn policy distributions that are
proportional to rewards and have limited exploration capabilities. In this
paper, we innovatively formulate the graph active learning problem as a
generative process, named GFlowGNN, which generates various samples through
sequential actions with probabilities precisely proportional to a predefined
reward function. Furthermore, we propose the concept of flow nodes and flow
features to efficiently model graphs as flows based on generative flow
networks, where the policy network is trained with specially designed rewards.
Extensive experiments on real datasets show that the proposed approach has good
exploration capability and transferability, outperforming various
state-of-the-art methods.
| [
{
"version": "v1",
"created": "Mon, 24 Apr 2023 10:47:08 GMT"
}
] | 2023-04-25T00:00:00 | [
[
"Li",
"Yinchuan",
""
],
[
"Li",
"Zhigang",
""
],
[
"Li",
"Wenqian",
""
],
[
"Shao",
"Yunfeng",
""
],
[
"Zheng",
"Yan",
""
],
[
"Hao",
"Jianye",
""
]
] | TITLE: Generative Flow Networks for Precise Reward-Oriented Active Learning on
Graphs
ABSTRACT: Many score-based active learning methods have been successfully applied to
graph-structured data, aiming to reduce the number of labels and achieve better
performance of graph neural networks based on predefined score functions.
However, these algorithms struggle to learn policy distributions that are
proportional to rewards and have limited exploration capabilities. In this
paper, we innovatively formulate the graph active learning problem as a
generative process, named GFlowGNN, which generates various samples through
sequential actions with probabilities precisely proportional to a predefined
reward function. Furthermore, we propose the concept of flow nodes and flow
features to efficiently model graphs as flows based on generative flow
networks, where the policy network is trained with specially designed rewards.
Extensive experiments on real datasets show that the proposed approach has good
exploration capability and transferability, outperforming various
state-of-the-art methods.
| no_new_dataset | 0.71027 |
2304.12000 | Xianghua Zeng | Xianghua Zeng, Hao Peng, Angsheng Li, Chunyang Liu, Lifang He, Philip
S. Yu | Hierarchical State Abstraction Based on Structural Information
Principles | null | null | null | null | cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | State abstraction optimizes decision-making by ignoring irrelevant
environmental information in reinforcement learning with rich observations.
Nevertheless, recent approaches focus on adequate representational capacities
resulting in essential information loss, affecting their performances on
challenging tasks. In this article, we propose a novel mathematical Structural
Information principles-based State Abstraction framework, namely SISA, from the
information-theoretic perspective. Specifically, an unsupervised, adaptive
hierarchical state clustering method without requiring manual assistance is
presented, and meanwhile, an optimal encoding tree is generated. On each
non-root tree node, a new aggregation function and condition structural entropy
are designed to achieve hierarchical state abstraction and compensate for
sampling-induced essential information loss in state abstraction. Empirical
evaluations on a visual gridworld domain and six continuous control benchmarks
demonstrate that, compared with five SOTA state abstraction approaches, SISA
significantly improves mean episode reward and sample efficiency up to 18.98
and 44.44%, respectively. Besides, we experimentally show that SISA is a
general framework that can be flexibly integrated with different
representation-learning objectives to improve their performances further.
| [
{
"version": "v1",
"created": "Mon, 24 Apr 2023 11:06:52 GMT"
}
] | 2023-04-25T00:00:00 | [
[
"Zeng",
"Xianghua",
""
],
[
"Peng",
"Hao",
""
],
[
"Li",
"Angsheng",
""
],
[
"Liu",
"Chunyang",
""
],
[
"He",
"Lifang",
""
],
[
"Yu",
"Philip S.",
""
]
] | TITLE: Hierarchical State Abstraction Based on Structural Information
Principles
ABSTRACT: State abstraction optimizes decision-making by ignoring irrelevant
environmental information in reinforcement learning with rich observations.
Nevertheless, recent approaches focus on adequate representational capacities
resulting in essential information loss, affecting their performances on
challenging tasks. In this article, we propose a novel mathematical Structural
Information principles-based State Abstraction framework, namely SISA, from the
information-theoretic perspective. Specifically, an unsupervised, adaptive
hierarchical state clustering method without requiring manual assistance is
presented, and meanwhile, an optimal encoding tree is generated. On each
non-root tree node, a new aggregation function and condition structural entropy
are designed to achieve hierarchical state abstraction and compensate for
sampling-induced essential information loss in state abstraction. Empirical
evaluations on a visual gridworld domain and six continuous control benchmarks
demonstrate that, compared with five SOTA state abstraction approaches, SISA
significantly improves mean episode reward and sample efficiency up to 18.98
and 44.44%, respectively. Besides, we experimentally show that SISA is a
general framework that can be flexibly integrated with different
representation-learning objectives to improve their performances further.
| no_new_dataset | 0.710239 |
2304.12026 | Haolan Zhan | Haolan Zhan and Zhuang Li and Yufei Wang and Linhao Luo and Tao Feng
and Xiaoxi Kang and Yuncheng Hua and Lizhen Qu and Lay-Ki Soon and Suraj
Sharma and Ingrid Zukerman and Zhaleh Semnani-Azad and Gholamreza Haffari | SocialDial: A Benchmark for Socially-Aware Dialogue Systems | Accepted by SIGIR 2023 | null | null | null | cs.CL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Dialogue systems have been widely applied in many scenarios and are now more
powerful and ubiquitous than ever before. With large neural models and massive
available data, current dialogue systems have access to more knowledge than any
people in their life. However, current dialogue systems still do not perform at
a human level. One major gap between conversational agents and humans lies in
their abilities to be aware of social norms. The development of socially-aware
dialogue systems is impeded due to the lack of resources. In this paper, we
present the first socially-aware dialogue corpus - SocialDial, based on Chinese
social culture. SocialDial consists of two parts: 1,563 multi-turn dialogues
between two human speakers with fine-grained labels, and 4,870 synthetic
conversations generated by ChatGPT. The human corpus covers five categories of
social norms, which have 14 sub-categories in total. Specifically, it contains
social factor annotations including social relation, context, social distance,
and social norms. However, collecting sufficient socially-aware dialogues is
costly. Thus, we harness the power of ChatGPT and devise an ontology-based
synthetic data generation framework. This framework is able to generate
synthetic data at scale. To ensure the quality of synthetic dialogues, we
design several mechanisms for quality control during data collection. Finally,
we evaluate our dataset using several pre-trained models, such as BERT and
RoBERTa. Comprehensive empirical results based on state-of-the-art neural
models demonstrate that modeling of social norms for dialogue systems is a
promising research direction. To the best of our knowledge, SocialDial is the
first socially-aware dialogue dataset that covers multiple social factors and
has fine-grained labels.
| [
{
"version": "v1",
"created": "Mon, 24 Apr 2023 11:55:22 GMT"
}
] | 2023-04-25T00:00:00 | [
[
"Zhan",
"Haolan",
""
],
[
"Li",
"Zhuang",
""
],
[
"Wang",
"Yufei",
""
],
[
"Luo",
"Linhao",
""
],
[
"Feng",
"Tao",
""
],
[
"Kang",
"Xiaoxi",
""
],
[
"Hua",
"Yuncheng",
""
],
[
"Qu",
"Lizhen",
""
],
[
"Soon",
"Lay-Ki",
""
],
[
"Sharma",
"Suraj",
""
],
[
"Zukerman",
"Ingrid",
""
],
[
"Semnani-Azad",
"Zhaleh",
""
],
[
"Haffari",
"Gholamreza",
""
]
] | TITLE: SocialDial: A Benchmark for Socially-Aware Dialogue Systems
ABSTRACT: Dialogue systems have been widely applied in many scenarios and are now more
powerful and ubiquitous than ever before. With large neural models and massive
available data, current dialogue systems have access to more knowledge than any
people in their life. However, current dialogue systems still do not perform at
a human level. One major gap between conversational agents and humans lies in
their abilities to be aware of social norms. The development of socially-aware
dialogue systems is impeded due to the lack of resources. In this paper, we
present the first socially-aware dialogue corpus - SocialDial, based on Chinese
social culture. SocialDial consists of two parts: 1,563 multi-turn dialogues
between two human speakers with fine-grained labels, and 4,870 synthetic
conversations generated by ChatGPT. The human corpus covers five categories of
social norms, which have 14 sub-categories in total. Specifically, it contains
social factor annotations including social relation, context, social distance,
and social norms. However, collecting sufficient socially-aware dialogues is
costly. Thus, we harness the power of ChatGPT and devise an ontology-based
synthetic data generation framework. This framework is able to generate
synthetic data at scale. To ensure the quality of synthetic dialogues, we
design several mechanisms for quality control during data collection. Finally,
we evaluate our dataset using several pre-trained models, such as BERT and
RoBERTa. Comprehensive empirical results based on state-of-the-art neural
models demonstrate that modeling of social norms for dialogue systems is a
promising research direction. To the best of our knowledge, SocialDial is the
first socially-aware dialogue dataset that covers multiple social factors and
has fine-grained labels.
| new_dataset | 0.721148 |
2304.12035 | Wanglong Lu | Wanglong Lu, Xianta Jiang, Xiaogang Jin, Yong-Liang Yang, Minglun
Gong, Tao Wang, Kaijie Shi, and Hanli Zhao | GRIG: Few-Shot Generative Residual Image Inpainting | There are 12 pages and 10 figures in this paper | null | null | null | cs.CV cs.MM | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Image inpainting is the task of filling in missing or masked region of an
image with semantically meaningful contents. Recent methods have shown
significant improvement in dealing with large-scale missing regions. However,
these methods usually require large training datasets to achieve satisfactory
results and there has been limited research into training these models on a
small number of samples. To address this, we present a novel few-shot
generative residual image inpainting method that produces high-quality
inpainting results. The core idea is to propose an iterative residual reasoning
method that incorporates Convolutional Neural Networks (CNNs) for feature
extraction and Transformers for global reasoning within generative adversarial
networks, along with image-level and patch-level discriminators. We also
propose a novel forgery-patch adversarial training strategy to create faithful
textures and detailed appearances. Extensive evaluations show that our method
outperforms previous methods on the few-shot image inpainting task, both
quantitatively and qualitatively.
| [
{
"version": "v1",
"created": "Mon, 24 Apr 2023 12:19:06 GMT"
}
] | 2023-04-25T00:00:00 | [
[
"Lu",
"Wanglong",
""
],
[
"Jiang",
"Xianta",
""
],
[
"Jin",
"Xiaogang",
""
],
[
"Yang",
"Yong-Liang",
""
],
[
"Gong",
"Minglun",
""
],
[
"Wang",
"Tao",
""
],
[
"Shi",
"Kaijie",
""
],
[
"Zhao",
"Hanli",
""
]
] | TITLE: GRIG: Few-Shot Generative Residual Image Inpainting
ABSTRACT: Image inpainting is the task of filling in missing or masked region of an
image with semantically meaningful contents. Recent methods have shown
significant improvement in dealing with large-scale missing regions. However,
these methods usually require large training datasets to achieve satisfactory
results and there has been limited research into training these models on a
small number of samples. To address this, we present a novel few-shot
generative residual image inpainting method that produces high-quality
inpainting results. The core idea is to propose an iterative residual reasoning
method that incorporates Convolutional Neural Networks (CNNs) for feature
extraction and Transformers for global reasoning within generative adversarial
networks, along with image-level and patch-level discriminators. We also
propose a novel forgery-patch adversarial training strategy to create faithful
textures and detailed appearances. Extensive evaluations show that our method
outperforms previous methods on the few-shot image inpainting task, both
quantitatively and qualitatively.
| no_new_dataset | 0.708572 |
2304.12051 | Andre Rochow | Andre Rochow, Max Schwarz, Michael Schreiber, Sven Behnke | VR Facial Animation for Immersive Telepresence Avatars | Published in IEEE/RSJ International Conference on Intelligent Robots
and Systems (IROS) 2022 | null | 10.1109/IROS47612.2022.9981892 | null | cs.CV cs.RO | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | VR Facial Animation is necessary in applications requiring clear view of the
face, even though a VR headset is worn. In our case, we aim to animate the face
of an operator who is controlling our robotic avatar system. We propose a
real-time capable pipeline with very fast adaptation for specific operators. In
a quick enrollment step, we capture a sequence of source images from the
operator without the VR headset which contain all the important
operator-specific appearance information. During inference, we then use the
operator keypoint information extracted from a mouth camera and two eye cameras
to estimate the target expression and head pose, to which we map the appearance
of a source still image. In order to enhance the mouth expression accuracy, we
dynamically select an auxiliary expression frame from the captured sequence.
This selection is done by learning to transform the current mouth keypoints
into the source camera space, where the alignment can be determined accurately.
We, furthermore, demonstrate an eye tracking pipeline that can be trained in
less than a minute, a time efficient way to train the whole pipeline given a
dataset that includes only complete faces, show exemplary results generated by
our method, and discuss performance at the ANA Avatar XPRIZE semifinals.
| [
{
"version": "v1",
"created": "Mon, 24 Apr 2023 12:43:51 GMT"
}
] | 2023-04-25T00:00:00 | [
[
"Rochow",
"Andre",
""
],
[
"Schwarz",
"Max",
""
],
[
"Schreiber",
"Michael",
""
],
[
"Behnke",
"Sven",
""
]
] | TITLE: VR Facial Animation for Immersive Telepresence Avatars
ABSTRACT: VR Facial Animation is necessary in applications requiring clear view of the
face, even though a VR headset is worn. In our case, we aim to animate the face
of an operator who is controlling our robotic avatar system. We propose a
real-time capable pipeline with very fast adaptation for specific operators. In
a quick enrollment step, we capture a sequence of source images from the
operator without the VR headset which contain all the important
operator-specific appearance information. During inference, we then use the
operator keypoint information extracted from a mouth camera and two eye cameras
to estimate the target expression and head pose, to which we map the appearance
of a source still image. In order to enhance the mouth expression accuracy, we
dynamically select an auxiliary expression frame from the captured sequence.
This selection is done by learning to transform the current mouth keypoints
into the source camera space, where the alignment can be determined accurately.
We, furthermore, demonstrate an eye tracking pipeline that can be trained in
less than a minute, a time efficient way to train the whole pipeline given a
dataset that includes only complete faces, show exemplary results generated by
our method, and discuss performance at the ANA Avatar XPRIZE semifinals.
| no_new_dataset | 0.684488 |
2304.12082 | Xiangming Gu | Xiangming Gu, Wei Zeng, Jianan Zhang, Longshen Ou, Ye Wang | Deep Audio-Visual Singing Voice Transcription based on Self-Supervised
Learning Models | null | null | null | null | cs.SD eess.AS | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Singing voice transcription converts recorded singing audio to musical
notation. Sound contamination (such as accompaniment) and lack of annotated
data make singing voice transcription an extremely difficult task. We take two
approaches to tackle the above challenges: 1) introducing multimodal learning
for singing voice transcription together with a new multimodal singing dataset,
N20EMv2, enhancing noise robustness by utilizing video information (lip
movements to predict the onset/offset of notes), and 2) adapting
self-supervised learning models from the speech domain to the singing voice
transcription task, significantly reducing annotated data requirements while
preserving pretrained features. We build a self-supervised learning based
audio-only singing voice transcription system, which not only outperforms
current state-of-the-art technologies as a strong baseline, but also
generalizes well to out-of-domain singing data. We then develop a
self-supervised learning based video-only singing voice transcription system
that detects note onsets and offsets with an accuracy of about 80\%. Finally,
based on the powerful acoustic and visual representations extracted by the
above two systems as well as the feature fusion design, we create an
audio-visual singing voice transcription system that improves the noise
robustness significantly under different acoustic environments compared to the
audio-only systems.
| [
{
"version": "v1",
"created": "Mon, 24 Apr 2023 13:25:22 GMT"
}
] | 2023-04-25T00:00:00 | [
[
"Gu",
"Xiangming",
""
],
[
"Zeng",
"Wei",
""
],
[
"Zhang",
"Jianan",
""
],
[
"Ou",
"Longshen",
""
],
[
"Wang",
"Ye",
""
]
] | TITLE: Deep Audio-Visual Singing Voice Transcription based on Self-Supervised
Learning Models
ABSTRACT: Singing voice transcription converts recorded singing audio to musical
notation. Sound contamination (such as accompaniment) and lack of annotated
data make singing voice transcription an extremely difficult task. We take two
approaches to tackle the above challenges: 1) introducing multimodal learning
for singing voice transcription together with a new multimodal singing dataset,
N20EMv2, enhancing noise robustness by utilizing video information (lip
movements to predict the onset/offset of notes), and 2) adapting
self-supervised learning models from the speech domain to the singing voice
transcription task, significantly reducing annotated data requirements while
preserving pretrained features. We build a self-supervised learning based
audio-only singing voice transcription system, which not only outperforms
current state-of-the-art technologies as a strong baseline, but also
generalizes well to out-of-domain singing data. We then develop a
self-supervised learning based video-only singing voice transcription system
that detects note onsets and offsets with an accuracy of about 80\%. Finally,
based on the powerful acoustic and visual representations extracted by the
above two systems as well as the feature fusion design, we create an
audio-visual singing voice transcription system that improves the noise
robustness significantly under different acoustic environments compared to the
audio-only systems.
| new_dataset | 0.714672 |
2304.12083 | Wei Yuan | Xuhui Ren, Wei Yuan, Tong Chen, Chaoqun Yang, Quoc Viet Hung Nguyen,
Hongzhi Yin | Joint Semantic and Structural Representation Learning for Enhancing User
Preference Modelling | null | null | null | null | cs.IR | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Knowledge graphs (KGs) have become important auxiliary information for
helping recommender systems obtain a good understanding of user preferences.
Despite recent advances in KG-based recommender systems, existing methods are
prone to suboptimal performance due to the following two drawbacks: 1) current
KG-based methods over-emphasize the heterogeneous structural information within
a KG and overlook the underlying semantics of its connections, hindering the
recommender from distilling the explicit user preferences; and 2) the inherent
incompleteness of a KG (i.e., missing facts, relations and entities) will
deteriorate the information extracted from KG and weaken the representation
learning of recommender systems.
To tackle the aforementioned problems, we investigate the potential of
jointly incorporating the structural and semantic information within a KG to
model user preferences in finer granularity. A new framework for KG-based
recommender systems, namely \textit{K}nowledge \textit{I}nfomax
\textit{R}ecommender \textit{S}ystem with \textit{C}ontrastive
\textit{L}earning (KIRS-CL) is proposed in this paper. Distinct from previous
KG-based approaches, KIRS-CL utilizes structural and connectivity information
with high-quality item embeddings learned by encoding KG triples with a
pre-trained language model. These well-trained entity representations enable
KIRS-CL to find the item to recommend via the preference connection between the
user and the item. Additionally, to improve the generalizability of our
framework, we introduce a contrastive warm-up learning strategy, making it
capable of dealing with both warm- and cold-start recommendation scenarios.
Extensive experiments on two real-world datasets demonstrate remarkable
improvements over state-of-the-art baselines.
| [
{
"version": "v1",
"created": "Mon, 24 Apr 2023 13:26:41 GMT"
}
] | 2023-04-25T00:00:00 | [
[
"Ren",
"Xuhui",
""
],
[
"Yuan",
"Wei",
""
],
[
"Chen",
"Tong",
""
],
[
"Yang",
"Chaoqun",
""
],
[
"Nguyen",
"Quoc Viet Hung",
""
],
[
"Yin",
"Hongzhi",
""
]
] | TITLE: Joint Semantic and Structural Representation Learning for Enhancing User
Preference Modelling
ABSTRACT: Knowledge graphs (KGs) have become important auxiliary information for
helping recommender systems obtain a good understanding of user preferences.
Despite recent advances in KG-based recommender systems, existing methods are
prone to suboptimal performance due to the following two drawbacks: 1) current
KG-based methods over-emphasize the heterogeneous structural information within
a KG and overlook the underlying semantics of its connections, hindering the
recommender from distilling the explicit user preferences; and 2) the inherent
incompleteness of a KG (i.e., missing facts, relations and entities) will
deteriorate the information extracted from KG and weaken the representation
learning of recommender systems.
To tackle the aforementioned problems, we investigate the potential of
jointly incorporating the structural and semantic information within a KG to
model user preferences in finer granularity. A new framework for KG-based
recommender systems, namely \textit{K}nowledge \textit{I}nfomax
\textit{R}ecommender \textit{S}ystem with \textit{C}ontrastive
\textit{L}earning (KIRS-CL) is proposed in this paper. Distinct from previous
KG-based approaches, KIRS-CL utilizes structural and connectivity information
with high-quality item embeddings learned by encoding KG triples with a
pre-trained language model. These well-trained entity representations enable
KIRS-CL to find the item to recommend via the preference connection between the
user and the item. Additionally, to improve the generalizability of our
framework, we introduce a contrastive warm-up learning strategy, making it
capable of dealing with both warm- and cold-start recommendation scenarios.
Extensive experiments on two real-world datasets demonstrate remarkable
improvements over state-of-the-art baselines.
| no_new_dataset | 0.706406 |
2304.12085 | Yunlong Xu | Yunlong Xu, Peizhen Yang, Zhengbin Tao | Dangoron: Network Construction on Large-scale Time Series Data across
Sliding Windows | null | null | null | null | physics.soc-ph cs.DB | http://creativecommons.org/licenses/by/4.0/ | Complex networks represent system dynamics through the interactions of a set
of anomalous time series. Consider the problem of computing correlations for
highly correlated pairs of time series across sliding windows. Efficiently
computing and updating the correlation matrix for user-defined sliding periods
and thresholds enables large-scale time series network dynamics analysis. We
introduce Dangoron, a framework for effectively identifying highly correlated
pairs of time series over sliding windows and computing their exact
correlation. By predicting dynamic correlation across sliding windows and
pruning unrelated time series, Dangoron is at least an order of magnitude
faster than a baseline. Additionally, we propose Tomborg, the first benchmark
for the problem of correlation matrix computation.
| [
{
"version": "v1",
"created": "Tue, 11 Apr 2023 23:54:41 GMT"
}
] | 2023-04-25T00:00:00 | [
[
"Xu",
"Yunlong",
""
],
[
"Yang",
"Peizhen",
""
],
[
"Tao",
"Zhengbin",
""
]
] | TITLE: Dangoron: Network Construction on Large-scale Time Series Data across
Sliding Windows
ABSTRACT: Complex networks represent system dynamics through the interactions of a set
of anomalous time series. Consider the problem of computing correlations for
highly correlated pairs of time series across sliding windows. Efficiently
computing and updating the correlation matrix for user-defined sliding periods
and thresholds enables large-scale time series network dynamics analysis. We
introduce Dangoron, a framework for effectively identifying highly correlated
pairs of time series over sliding windows and computing their exact
correlation. By predicting dynamic correlation across sliding windows and
pruning unrelated time series, Dangoron is at least an order of magnitude
faster than a baseline. Additionally, we propose Tomborg, the first benchmark
for the problem of correlation matrix computation.
| no_new_dataset | 0.709787 |
2304.12115 | Bal\'azs Pej\'o | Balazs Pejo and Nikolett Kapui | SQLi Detection with ML: A data-source perspective | Extended version of an accepted paper at SECRYPT 2023 | null | null | null | cs.CR cs.LG | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Almost 50 years after the invention of SQL, injection attacks are still
top-tier vulnerabilities of today's ICT systems. Consequently, SQLi detection
is still an active area of research, where the most recent works incorporate
machine learning techniques into the proposed solutions. In this work, we
highlight the shortcomings of the previous ML-based results focusing on four
aspects: the evaluation methods, the optimization of the model parameters, the
distribution of utilized datasets, and the feature selection. Since no single
work explored all of these aspects satisfactorily, we fill this gap and provide
an in-depth and comprehensive empirical analysis. Moreover, we cross-validate
the trained models by using data from other distributions. This aspect of ML
models (trained for SQLi detection) was never studied. Yet, the sensitivity of
the model's performance to this is crucial for any real-life deployment.
Finally, we validate our findings on a real-world industrial SQLi dataset.
| [
{
"version": "v1",
"created": "Mon, 24 Apr 2023 14:17:34 GMT"
}
] | 2023-04-25T00:00:00 | [
[
"Pejo",
"Balazs",
""
],
[
"Kapui",
"Nikolett",
""
]
] | TITLE: SQLi Detection with ML: A data-source perspective
ABSTRACT: Almost 50 years after the invention of SQL, injection attacks are still
top-tier vulnerabilities of today's ICT systems. Consequently, SQLi detection
is still an active area of research, where the most recent works incorporate
machine learning techniques into the proposed solutions. In this work, we
highlight the shortcomings of the previous ML-based results focusing on four
aspects: the evaluation methods, the optimization of the model parameters, the
distribution of utilized datasets, and the feature selection. Since no single
work explored all of these aspects satisfactorily, we fill this gap and provide
an in-depth and comprehensive empirical analysis. Moreover, we cross-validate
the trained models by using data from other distributions. This aspect of ML
models (trained for SQLi detection) was never studied. Yet, the sensitivity of
the model's performance to this is crucial for any real-life deployment.
Finally, we validate our findings on a real-world industrial SQLi dataset.
| no_new_dataset | 0.709233 |
2304.12122 | Manuel Schwonberg | Manuel Schwonberg, Fadoua El Bouazati, Nico M. Schmidt, Hanno
Gottschalk | Augmentation-based Domain Generalization for Semantic Segmentation | Accepted at Intelligent Vehicles Symposium 2023 (IV 2023)
Autonomy@Scale Workshop | null | null | null | cs.CV cs.AI cs.LG cs.RO | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Unsupervised Domain Adaptation (UDA) and domain generalization (DG) are two
research areas that aim to tackle the lack of generalization of Deep Neural
Networks (DNNs) towards unseen domains. While UDA methods have access to
unlabeled target images, domain generalization does not involve any target data
and only learns generalized features from a source domain. Image-style
randomization or augmentation is a popular approach to improve network
generalization without access to the target domain. Complex methods are often
proposed that disregard the potential of simple image augmentations for
out-of-domain generalization. For this reason, we systematically study the in-
and out-of-domain generalization capabilities of simple, rule-based image
augmentations like blur, noise, color jitter and many more. Based on a full
factorial design of experiment design we provide a systematic statistical
evaluation of augmentations and their interactions. Our analysis provides both,
expected and unexpected, outcomes. Expected, because our experiments confirm
the common scientific standard that combination of multiple different
augmentations out-performs single augmentations. Unexpected, because combined
augmentations perform competitive to state-of-the-art domain generalization
approaches, while being significantly simpler and without training overhead. On
the challenging synthetic-to-real domain shift between Synthia and Cityscapes
we reach 39.5% mIoU compared to 40.9% mIoU of the best previous work. When
additionally employing the recent vision transformer architecture DAFormer we
outperform these benchmarks with a performance of 44.2% mIoU
| [
{
"version": "v1",
"created": "Mon, 24 Apr 2023 14:26:53 GMT"
}
] | 2023-04-25T00:00:00 | [
[
"Schwonberg",
"Manuel",
""
],
[
"Bouazati",
"Fadoua El",
""
],
[
"Schmidt",
"Nico M.",
""
],
[
"Gottschalk",
"Hanno",
""
]
] | TITLE: Augmentation-based Domain Generalization for Semantic Segmentation
ABSTRACT: Unsupervised Domain Adaptation (UDA) and domain generalization (DG) are two
research areas that aim to tackle the lack of generalization of Deep Neural
Networks (DNNs) towards unseen domains. While UDA methods have access to
unlabeled target images, domain generalization does not involve any target data
and only learns generalized features from a source domain. Image-style
randomization or augmentation is a popular approach to improve network
generalization without access to the target domain. Complex methods are often
proposed that disregard the potential of simple image augmentations for
out-of-domain generalization. For this reason, we systematically study the in-
and out-of-domain generalization capabilities of simple, rule-based image
augmentations like blur, noise, color jitter and many more. Based on a full
factorial design of experiment design we provide a systematic statistical
evaluation of augmentations and their interactions. Our analysis provides both,
expected and unexpected, outcomes. Expected, because our experiments confirm
the common scientific standard that combination of multiple different
augmentations out-performs single augmentations. Unexpected, because combined
augmentations perform competitive to state-of-the-art domain generalization
approaches, while being significantly simpler and without training overhead. On
the challenging synthetic-to-real domain shift between Synthia and Cityscapes
we reach 39.5% mIoU compared to 40.9% mIoU of the best previous work. When
additionally employing the recent vision transformer architecture DAFormer we
outperform these benchmarks with a performance of 44.2% mIoU
| no_new_dataset | 0.709403 |
2304.12139 | Jimmy Lin | Xueguang Ma, Tommaso Teofili, Jimmy Lin | Anserini Gets Dense Retrieval: Integration of Lucene's HNSW Indexes | null | null | null | null | cs.IR | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Anserini is a Lucene-based toolkit for reproducible information retrieval
research in Java that has been gaining traction in the community. It provides
retrieval capabilities for both "traditional" bag-of-words retrieval models
such as BM25 as well as retrieval using learned sparse representations such as
SPLADE. With Pyserini, which provides a Python interface to Anserini, users
gain access to both sparse and dense retrieval models, as Pyserini implements
bindings to the Faiss vector search library alongside Lucene inverted indexes
in a uniform, consistent interface. Nevertheless, hybrid fusion techniques that
integrate sparse and dense retrieval models need to stitch together results
from two completely different "software stacks", which creates unnecessary
complexities and inefficiencies. However, the introduction of HNSW indexes for
dense vector search in Lucene promises the integration of both dense and sparse
retrieval within a single software framework. We explore exactly this
integration in the context of Anserini. Experiments on the MS MARCO passage and
BEIR datasets show that our Anserini HNSW integration supports (reasonably)
effective and (reasonably) efficient approximate nearest neighbor search for
dense retrieval models, using only Lucene.
| [
{
"version": "v1",
"created": "Mon, 24 Apr 2023 14:44:27 GMT"
}
] | 2023-04-25T00:00:00 | [
[
"Ma",
"Xueguang",
""
],
[
"Teofili",
"Tommaso",
""
],
[
"Lin",
"Jimmy",
""
]
] | TITLE: Anserini Gets Dense Retrieval: Integration of Lucene's HNSW Indexes
ABSTRACT: Anserini is a Lucene-based toolkit for reproducible information retrieval
research in Java that has been gaining traction in the community. It provides
retrieval capabilities for both "traditional" bag-of-words retrieval models
such as BM25 as well as retrieval using learned sparse representations such as
SPLADE. With Pyserini, which provides a Python interface to Anserini, users
gain access to both sparse and dense retrieval models, as Pyserini implements
bindings to the Faiss vector search library alongside Lucene inverted indexes
in a uniform, consistent interface. Nevertheless, hybrid fusion techniques that
integrate sparse and dense retrieval models need to stitch together results
from two completely different "software stacks", which creates unnecessary
complexities and inefficiencies. However, the introduction of HNSW indexes for
dense vector search in Lucene promises the integration of both dense and sparse
retrieval within a single software framework. We explore exactly this
integration in the context of Anserini. Experiments on the MS MARCO passage and
BEIR datasets show that our Anserini HNSW integration supports (reasonably)
effective and (reasonably) efficient approximate nearest neighbor search for
dense retrieval models, using only Lucene.
| no_new_dataset | 0.707556 |
2304.12160 | Anurag Arnab | Alexey Gritsenko, Xuehan Xiong, Josip Djolonga, Mostafa Dehghani, Chen
Sun, Mario Lu\v{c}i\'c, Cordelia Schmid, Anurag Arnab | End-to-End Spatio-Temporal Action Localisation with Video Transformers | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The most performant spatio-temporal action localisation models use external
person proposals and complex external memory banks. We propose a fully
end-to-end, purely-transformer based model that directly ingests an input
video, and outputs tubelets -- a sequence of bounding boxes and the action
classes at each frame. Our flexible model can be trained with either sparse
bounding-box supervision on individual frames, or full tubelet annotations. And
in both cases, it predicts coherent tubelets as the output. Moreover, our
end-to-end model requires no additional pre-processing in the form of
proposals, or post-processing in terms of non-maximal suppression. We perform
extensive ablation experiments, and significantly advance the state-of-the-art
results on four different spatio-temporal action localisation benchmarks with
both sparse keyframes and full tubelet annotations.
| [
{
"version": "v1",
"created": "Mon, 24 Apr 2023 15:14:01 GMT"
}
] | 2023-04-25T00:00:00 | [
[
"Gritsenko",
"Alexey",
""
],
[
"Xiong",
"Xuehan",
""
],
[
"Djolonga",
"Josip",
""
],
[
"Dehghani",
"Mostafa",
""
],
[
"Sun",
"Chen",
""
],
[
"Lučić",
"Mario",
""
],
[
"Schmid",
"Cordelia",
""
],
[
"Arnab",
"Anurag",
""
]
] | TITLE: End-to-End Spatio-Temporal Action Localisation with Video Transformers
ABSTRACT: The most performant spatio-temporal action localisation models use external
person proposals and complex external memory banks. We propose a fully
end-to-end, purely-transformer based model that directly ingests an input
video, and outputs tubelets -- a sequence of bounding boxes and the action
classes at each frame. Our flexible model can be trained with either sparse
bounding-box supervision on individual frames, or full tubelet annotations. And
in both cases, it predicts coherent tubelets as the output. Moreover, our
end-to-end model requires no additional pre-processing in the form of
proposals, or post-processing in terms of non-maximal suppression. We perform
extensive ablation experiments, and significantly advance the state-of-the-art
results on four different spatio-temporal action localisation benchmarks with
both sparse keyframes and full tubelet annotations.
| no_new_dataset | 0.710672 |
2304.12161 | Berkan Demirel | Berkan Demirel, Orhun Bu\u{g}ra Baran, Ramazan Gokberk Cinbis | Meta-tuning Loss Functions and Data Augmentation for Few-shot Object
Detection | To appear at IEEE/CVF CVPR 2023 | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Few-shot object detection, the problem of modelling novel object detection
categories with few training instances, is an emerging topic in the area of
few-shot learning and object detection. Contemporary techniques can be divided
into two groups: fine-tuning based and meta-learning based approaches. While
meta-learning approaches aim to learn dedicated meta-models for mapping samples
to novel class models, fine-tuning approaches tackle few-shot detection in a
simpler manner, by adapting the detection model to novel classes through
gradient based optimization. Despite their simplicity, fine-tuning based
approaches typically yield competitive detection results. Based on this
observation, we focus on the role of loss functions and augmentations as the
force driving the fine-tuning process, and propose to tune their dynamics
through meta-learning principles. The proposed training scheme, therefore,
allows learning inductive biases that can boost few-shot detection, while
keeping the advantages of fine-tuning based approaches. In addition, the
proposed approach yields interpretable loss functions, as opposed to highly
parametric and complex few-shot meta-models. The experimental results highlight
the merits of the proposed scheme, with significant improvements over the
strong fine-tuning based few-shot detection baselines on benchmark Pascal VOC
and MS-COCO datasets, in terms of both standard and generalized few-shot
performance metrics.
| [
{
"version": "v1",
"created": "Mon, 24 Apr 2023 15:14:16 GMT"
}
] | 2023-04-25T00:00:00 | [
[
"Demirel",
"Berkan",
""
],
[
"Baran",
"Orhun Buğra",
""
],
[
"Cinbis",
"Ramazan Gokberk",
""
]
] | TITLE: Meta-tuning Loss Functions and Data Augmentation for Few-shot Object
Detection
ABSTRACT: Few-shot object detection, the problem of modelling novel object detection
categories with few training instances, is an emerging topic in the area of
few-shot learning and object detection. Contemporary techniques can be divided
into two groups: fine-tuning based and meta-learning based approaches. While
meta-learning approaches aim to learn dedicated meta-models for mapping samples
to novel class models, fine-tuning approaches tackle few-shot detection in a
simpler manner, by adapting the detection model to novel classes through
gradient based optimization. Despite their simplicity, fine-tuning based
approaches typically yield competitive detection results. Based on this
observation, we focus on the role of loss functions and augmentations as the
force driving the fine-tuning process, and propose to tune their dynamics
through meta-learning principles. The proposed training scheme, therefore,
allows learning inductive biases that can boost few-shot detection, while
keeping the advantages of fine-tuning based approaches. In addition, the
proposed approach yields interpretable loss functions, as opposed to highly
parametric and complex few-shot meta-models. The experimental results highlight
the merits of the proposed scheme, with significant improvements over the
strong fine-tuning based few-shot detection baselines on benchmark Pascal VOC
and MS-COCO datasets, in terms of both standard and generalized few-shot
performance metrics.
| no_new_dataset | 0.709422 |
2304.12184 | Huabing Lu | Zhaoyuan Shi, Huabing Lu, Xianzhong Xie, Helin Yang, Chongwen Huang,
Jun Cai, and Zhiguo Ding | Active RIS-aided EH-NOMA Networks: A Deep Reinforcement Learning
Approach | null | null | null | null | eess.SP cs.AI cs.IT cs.LG math.IT | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | An active reconfigurable intelligent surface (RIS)-aided multi-user downlink
communication system is investigated, where non-orthogonal multiple access
(NOMA) is employed to improve spectral efficiency, and the active RIS is
powered by energy harvesting (EH). The problem of joint control of the RIS's
amplification matrix and phase shift matrix is formulated to maximize the
communication success ratio with considering the quality of service (QoS)
requirements of users, dynamic communication state, and dynamic available
energy of RIS. To tackle this non-convex problem, a cascaded deep learning
algorithm namely long short-term memory-deep deterministic policy gradient
(LSTM-DDPG) is designed. First, an advanced LSTM based algorithm is developed
to predict users' dynamic communication state. Then, based on the prediction
results, a DDPG based algorithm is proposed to joint control the amplification
matrix and phase shift matrix of the RIS. Finally, simulation results verify
the accuracy of the prediction of the proposed LSTM algorithm, and demonstrate
that the LSTM-DDPG algorithm has a significant advantage over other benchmark
algorithms in terms of communication success ratio performance.
| [
{
"version": "v1",
"created": "Tue, 11 Apr 2023 13:16:28 GMT"
}
] | 2023-04-25T00:00:00 | [
[
"Shi",
"Zhaoyuan",
""
],
[
"Lu",
"Huabing",
""
],
[
"Xie",
"Xianzhong",
""
],
[
"Yang",
"Helin",
""
],
[
"Huang",
"Chongwen",
""
],
[
"Cai",
"Jun",
""
],
[
"Ding",
"Zhiguo",
""
]
] | TITLE: Active RIS-aided EH-NOMA Networks: A Deep Reinforcement Learning
Approach
ABSTRACT: An active reconfigurable intelligent surface (RIS)-aided multi-user downlink
communication system is investigated, where non-orthogonal multiple access
(NOMA) is employed to improve spectral efficiency, and the active RIS is
powered by energy harvesting (EH). The problem of joint control of the RIS's
amplification matrix and phase shift matrix is formulated to maximize the
communication success ratio with considering the quality of service (QoS)
requirements of users, dynamic communication state, and dynamic available
energy of RIS. To tackle this non-convex problem, a cascaded deep learning
algorithm namely long short-term memory-deep deterministic policy gradient
(LSTM-DDPG) is designed. First, an advanced LSTM based algorithm is developed
to predict users' dynamic communication state. Then, based on the prediction
results, a DDPG based algorithm is proposed to joint control the amplification
matrix and phase shift matrix of the RIS. Finally, simulation results verify
the accuracy of the prediction of the proposed LSTM algorithm, and demonstrate
that the LSTM-DDPG algorithm has a significant advantage over other benchmark
algorithms in terms of communication success ratio performance.
| no_new_dataset | 0.709239 |
2304.12185 | Chih-Hsun Lin | Chih-Hsun Lin, Chia-Yi Hsu, Chia-Mu Yu, Yang Cao, Chun-Ying Huang | DPAF: Image Synthesis via Differentially Private Aggregation in Forward
Phase | null | null | null | null | cs.CV cs.CR cs.LG | http://creativecommons.org/licenses/by/4.0/ | Differentially private synthetic data is a promising alternative for
sensitive data release. Many differentially private generative models have been
proposed in the literature. Unfortunately, they all suffer from the low utility
of the synthetic data, particularly for images of high resolutions. Here, we
propose DPAF, an effective differentially private generative model for
high-dimensional image synthesis. Different from the prior private stochastic
gradient descent-based methods that add Gaussian noises in the backward phase
during the model training, DPAF adds a differentially private feature
aggregation in the forward phase, bringing advantages, including the reduction
of information loss in gradient clipping and low sensitivity for the
aggregation. Moreover, as an improper batch size has an adverse impact on the
utility of synthetic data, DPAF also tackles the problem of setting a proper
batch size by proposing a novel training strategy that asymmetrically trains
different parts of the discriminator. We extensively evaluate different methods
on multiple image datasets (up to images of 128x128 resolution) to demonstrate
the performance of DPAF.
| [
{
"version": "v1",
"created": "Thu, 20 Apr 2023 16:32:02 GMT"
}
] | 2023-04-25T00:00:00 | [
[
"Lin",
"Chih-Hsun",
""
],
[
"Hsu",
"Chia-Yi",
""
],
[
"Yu",
"Chia-Mu",
""
],
[
"Cao",
"Yang",
""
],
[
"Huang",
"Chun-Ying",
""
]
] | TITLE: DPAF: Image Synthesis via Differentially Private Aggregation in Forward
Phase
ABSTRACT: Differentially private synthetic data is a promising alternative for
sensitive data release. Many differentially private generative models have been
proposed in the literature. Unfortunately, they all suffer from the low utility
of the synthetic data, particularly for images of high resolutions. Here, we
propose DPAF, an effective differentially private generative model for
high-dimensional image synthesis. Different from the prior private stochastic
gradient descent-based methods that add Gaussian noises in the backward phase
during the model training, DPAF adds a differentially private feature
aggregation in the forward phase, bringing advantages, including the reduction
of information loss in gradient clipping and low sensitivity for the
aggregation. Moreover, as an improper batch size has an adverse impact on the
utility of synthetic data, DPAF also tackles the problem of setting a proper
batch size by proposing a novel training strategy that asymmetrically trains
different parts of the discriminator. We extensively evaluate different methods
on multiple image datasets (up to images of 128x128 resolution) to demonstrate
the performance of DPAF.
| no_new_dataset | 0.7114 |
2304.12194 | Deng Shuchao | Shuchao Deng, Yanan Sun, and Edgar Galvan | Neural Architecture Search Using Genetic Algorithm for Facial Expression
Recognition | null | null | null | null | cs.CV cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Facial expression is one of the most powerful, natural, and universal signals
for human beings to express emotional states and intentions. Thus, it is
evident the importance of correct and innovative facial expression recognition
(FER) approaches in Artificial Intelligence. The current common practice for
FER is to correctly design convolutional neural networks' architectures (CNNs)
using human expertise. However, finding a well-performing architecture is often
a very tedious and error-prone process for deep learning researchers. Neural
architecture search (NAS) is an area of growing interest as demonstrated by the
large number of scientific works published in recent years thanks to the
impressive results achieved in recent years. We propose a genetic algorithm
approach that uses an ingenious encoding-decoding mechanism that allows to
automatically evolve CNNs on FER tasks attaining high accuracy classification
rates. The experimental results demonstrate that the proposed algorithm
achieves the best-known results on the CK+ and FERG datasets as well as
competitive results on the JAFFE dataset.
| [
{
"version": "v1",
"created": "Wed, 12 Apr 2023 16:36:07 GMT"
}
] | 2023-04-25T00:00:00 | [
[
"Deng",
"Shuchao",
""
],
[
"Sun",
"Yanan",
""
],
[
"Galvan",
"Edgar",
""
]
] | TITLE: Neural Architecture Search Using Genetic Algorithm for Facial Expression
Recognition
ABSTRACT: Facial expression is one of the most powerful, natural, and universal signals
for human beings to express emotional states and intentions. Thus, it is
evident the importance of correct and innovative facial expression recognition
(FER) approaches in Artificial Intelligence. The current common practice for
FER is to correctly design convolutional neural networks' architectures (CNNs)
using human expertise. However, finding a well-performing architecture is often
a very tedious and error-prone process for deep learning researchers. Neural
architecture search (NAS) is an area of growing interest as demonstrated by the
large number of scientific works published in recent years thanks to the
impressive results achieved in recent years. We propose a genetic algorithm
approach that uses an ingenious encoding-decoding mechanism that allows to
automatically evolve CNNs on FER tasks attaining high accuracy classification
rates. The experimental results demonstrate that the proposed algorithm
achieves the best-known results on the CK+ and FERG datasets as well as
competitive results on the JAFFE dataset.
| no_new_dataset | 0.709818 |
2304.12202 | Ilias Chalkidis | Ilias Chalkidis | ChatGPT may Pass the Bar Exam soon, but has a Long Way to Go for the
LexGLUE benchmark | Working paper | null | null | null | cs.CL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Following the hype around OpenAI's ChatGPT conversational agent, the last
straw in the recent development of Large Language Models (LLMs) that
demonstrate emergent unprecedented zero-shot capabilities, we audit the latest
OpenAI's GPT-3.5 model, `gpt-3.5-turbo', the first available ChatGPT model, in
the LexGLUE benchmark in a zero-shot fashion providing examples in a templated
instruction-following format. The results indicate that ChatGPT achieves an
average micro-F1 score of 47.6% across LexGLUE tasks, surpassing the baseline
guessing rates. Notably, the model performs exceptionally well in some
datasets, achieving micro-F1 scores of 62.8% and 70.2% in the ECtHR B and
LEDGAR datasets, respectively. The code base and model predictions are
available for review on https://github.com/coastalcph/zeroshot_lexglue.
| [
{
"version": "v1",
"created": "Thu, 9 Mar 2023 16:42:29 GMT"
}
] | 2023-04-25T00:00:00 | [
[
"Chalkidis",
"Ilias",
""
]
] | TITLE: ChatGPT may Pass the Bar Exam soon, but has a Long Way to Go for the
LexGLUE benchmark
ABSTRACT: Following the hype around OpenAI's ChatGPT conversational agent, the last
straw in the recent development of Large Language Models (LLMs) that
demonstrate emergent unprecedented zero-shot capabilities, we audit the latest
OpenAI's GPT-3.5 model, `gpt-3.5-turbo', the first available ChatGPT model, in
the LexGLUE benchmark in a zero-shot fashion providing examples in a templated
instruction-following format. The results indicate that ChatGPT achieves an
average micro-F1 score of 47.6% across LexGLUE tasks, surpassing the baseline
guessing rates. Notably, the model performs exceptionally well in some
datasets, achieving micro-F1 scores of 62.8% and 70.2% in the ECtHR B and
LEDGAR datasets, respectively. The code base and model predictions are
available for review on https://github.com/coastalcph/zeroshot_lexglue.
| no_new_dataset | 0.709372 |
2304.12204 | Dong Won Lee | Dong Won Lee, Yubin Kim, Rosalind Picard, Cynthia Breazeal, Hae Won
Park | Multipar-T: Multiparty-Transformer for Capturing Contingent Behaviors in
Group Conversations | 7 pages, 4 figures, IJCAI | null | null | null | cs.CV cs.AI cs.LG | http://creativecommons.org/licenses/by/4.0/ | As we move closer to real-world AI systems, AI agents must be able to deal
with multiparty (group) conversations. Recognizing and interpreting multiparty
behaviors is challenging, as the system must recognize individual behavioral
cues, deal with the complexity of multiple streams of data from multiple
people, and recognize the subtle contingent social exchanges that take place
amongst group members. To tackle this challenge, we propose the
Multiparty-Transformer (Multipar-T), a transformer model for multiparty
behavior modeling. The core component of our proposed approach is the
Crossperson Attention, which is specifically designed to detect contingent
behavior between pairs of people. We verify the effectiveness of Multipar-T on
a publicly available video-based group engagement detection benchmark, where it
outperforms state-of-the-art approaches in average F-1 scores by 5.2% and
individual class F-1 scores by up to 10.0%. Through qualitative analysis, we
show that our Crossperson Attention module is able to discover contingent
behavior.
| [
{
"version": "v1",
"created": "Wed, 19 Apr 2023 20:23:11 GMT"
}
] | 2023-04-25T00:00:00 | [
[
"Lee",
"Dong Won",
""
],
[
"Kim",
"Yubin",
""
],
[
"Picard",
"Rosalind",
""
],
[
"Breazeal",
"Cynthia",
""
],
[
"Park",
"Hae Won",
""
]
] | TITLE: Multipar-T: Multiparty-Transformer for Capturing Contingent Behaviors in
Group Conversations
ABSTRACT: As we move closer to real-world AI systems, AI agents must be able to deal
with multiparty (group) conversations. Recognizing and interpreting multiparty
behaviors is challenging, as the system must recognize individual behavioral
cues, deal with the complexity of multiple streams of data from multiple
people, and recognize the subtle contingent social exchanges that take place
amongst group members. To tackle this challenge, we propose the
Multiparty-Transformer (Multipar-T), a transformer model for multiparty
behavior modeling. The core component of our proposed approach is the
Crossperson Attention, which is specifically designed to detect contingent
behavior between pairs of people. We verify the effectiveness of Multipar-T on
a publicly available video-based group engagement detection benchmark, where it
outperforms state-of-the-art approaches in average F-1 scores by 5.2% and
individual class F-1 scores by up to 10.0%. Through qualitative analysis, we
show that our Crossperson Attention module is able to discover contingent
behavior.
| no_new_dataset | 0.704402 |
2304.12239 | Guojiang Zhao | Zhifeng Gao, Xiaohong Ji, Guojiang Zhao, Hongshuai Wang, Hang Zheng,
Guolin Ke, Linfeng Zhang | Uni-QSAR: an Auto-ML Tool for Molecular Property Prediction | null | null | null | null | q-bio.BM cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Recently deep learning based quantitative structure-activity relationship
(QSAR) models has shown surpassing performance than traditional methods for
property prediction tasks in drug discovery. However, most DL based QSAR models
are restricted to limited labeled data to achieve better performance, and also
are sensitive to model scale and hyper-parameters. In this paper, we propose
Uni-QSAR, a powerful Auto-ML tool for molecule property prediction tasks.
Uni-QSAR combines molecular representation learning (MRL) of 1D sequential
tokens, 2D topology graphs, and 3D conformers with pretraining models to
leverage rich representation from large-scale unlabeled data. Without any
manual fine-tuning or model selection, Uni-QSAR outperforms SOTA in 21/22 tasks
of the Therapeutic Data Commons (TDC) benchmark under designed parallel
workflow, with an average performance improvement of 6.09\%. Furthermore, we
demonstrate the practical usefulness of Uni-QSAR in drug discovery domains.
| [
{
"version": "v1",
"created": "Mon, 24 Apr 2023 16:29:08 GMT"
}
] | 2023-04-25T00:00:00 | [
[
"Gao",
"Zhifeng",
""
],
[
"Ji",
"Xiaohong",
""
],
[
"Zhao",
"Guojiang",
""
],
[
"Wang",
"Hongshuai",
""
],
[
"Zheng",
"Hang",
""
],
[
"Ke",
"Guolin",
""
],
[
"Zhang",
"Linfeng",
""
]
] | TITLE: Uni-QSAR: an Auto-ML Tool for Molecular Property Prediction
ABSTRACT: Recently deep learning based quantitative structure-activity relationship
(QSAR) models has shown surpassing performance than traditional methods for
property prediction tasks in drug discovery. However, most DL based QSAR models
are restricted to limited labeled data to achieve better performance, and also
are sensitive to model scale and hyper-parameters. In this paper, we propose
Uni-QSAR, a powerful Auto-ML tool for molecule property prediction tasks.
Uni-QSAR combines molecular representation learning (MRL) of 1D sequential
tokens, 2D topology graphs, and 3D conformers with pretraining models to
leverage rich representation from large-scale unlabeled data. Without any
manual fine-tuning or model selection, Uni-QSAR outperforms SOTA in 21/22 tasks
of the Therapeutic Data Commons (TDC) benchmark under designed parallel
workflow, with an average performance improvement of 6.09\%. Furthermore, we
demonstrate the practical usefulness of Uni-QSAR in drug discovery domains.
| no_new_dataset | 0.708389 |
2304.12251 | \'Angel L\'opez-Oriona | \'Angel L\'opez Oriona and Jos\'e Antonio Vilar Fern\'andez | Ordinal time series analysis with the R package otsfeatures | null | null | null | null | stat.ML cs.LG | http://creativecommons.org/licenses/by/4.0/ | The 21st century has witnessed a growing interest in the analysis of time
series data. Whereas most of the literature on the topic deals with real-valued
time series, ordinal time series have typically received much less attention.
However, the development of specific analytical tools for the latter objects
has substantially increased in recent years. The R package otsfeatures attempts
to provide a set of simple functions for analyzing ordinal time series. In
particular, several commands allowing the extraction of well-known statistical
features and the execution of inferential tasks are available for the user. The
output of several functions can be employed to perform traditional machine
learning tasks including clustering, classification or outlier detection.
otsfeatures also incorporates two datasets of financial time series which were
used in the literature for clustering purposes, as well as three interesting
synthetic databases. The main properties of the package are described and its
use is illustrated through several examples. Researchers from a broad variety
of disciplines could benefit from the powerful tools provided by otsfeatures.
| [
{
"version": "v1",
"created": "Mon, 24 Apr 2023 16:40:27 GMT"
}
] | 2023-04-25T00:00:00 | [
[
"Oriona",
"Ángel López",
""
],
[
"Fernández",
"José Antonio Vilar",
""
]
] | TITLE: Ordinal time series analysis with the R package otsfeatures
ABSTRACT: The 21st century has witnessed a growing interest in the analysis of time
series data. Whereas most of the literature on the topic deals with real-valued
time series, ordinal time series have typically received much less attention.
However, the development of specific analytical tools for the latter objects
has substantially increased in recent years. The R package otsfeatures attempts
to provide a set of simple functions for analyzing ordinal time series. In
particular, several commands allowing the extraction of well-known statistical
features and the execution of inferential tasks are available for the user. The
output of several functions can be employed to perform traditional machine
learning tasks including clustering, classification or outlier detection.
otsfeatures also incorporates two datasets of financial time series which were
used in the literature for clustering purposes, as well as three interesting
synthetic databases. The main properties of the package are described and its
use is illustrated through several examples. Researchers from a broad variety
of disciplines could benefit from the powerful tools provided by otsfeatures.
| no_new_dataset | 0.707177 |
2304.12257 | Pablo Alonso-Jim\'enez | Pablo Alonso-Jim\'enez, Xavier Favory, Hadrien Foroughmand, Grigoris
Bourdalas, Xavier Serra, Thomas Lidy, Dmitry Bogdanov | Pre-Training Strategies Using Contrastive Learning and Playlist
Information for Music Classification and Similarity | Accepted at the 2023 International Conference on Acoustics, Speech,
and Signal Processing (ICASSP'23) | null | null | null | cs.SD eess.AS | http://creativecommons.org/licenses/by-nc-sa/4.0/ | In this work, we investigate an approach that relies on contrastive learning
and music metadata as a weak source of supervision to train music
representation models. Recent studies show that contrastive learning can be
used with editorial metadata (e.g., artist or album name) to learn audio
representations that are useful for different classification tasks. In this
paper, we extend this idea to using playlist data as a source of music
similarity information and investigate three approaches to generate anchor and
positive track pairs. We evaluate these approaches by fine-tuning the
pre-trained models for music multi-label classification tasks (genre, mood, and
instrument tagging) and music similarity. We find that creating anchor and
positive track pairs by relying on co-occurrences in playlists provides better
music similarity and competitive classification results compared to choosing
tracks from the same artist as in previous works. Additionally, our best
pre-training approach based on playlists provides superior classification
performance for most datasets.
| [
{
"version": "v1",
"created": "Mon, 24 Apr 2023 16:49:58 GMT"
}
] | 2023-04-25T00:00:00 | [
[
"Alonso-Jiménez",
"Pablo",
""
],
[
"Favory",
"Xavier",
""
],
[
"Foroughmand",
"Hadrien",
""
],
[
"Bourdalas",
"Grigoris",
""
],
[
"Serra",
"Xavier",
""
],
[
"Lidy",
"Thomas",
""
],
[
"Bogdanov",
"Dmitry",
""
]
] | TITLE: Pre-Training Strategies Using Contrastive Learning and Playlist
Information for Music Classification and Similarity
ABSTRACT: In this work, we investigate an approach that relies on contrastive learning
and music metadata as a weak source of supervision to train music
representation models. Recent studies show that contrastive learning can be
used with editorial metadata (e.g., artist or album name) to learn audio
representations that are useful for different classification tasks. In this
paper, we extend this idea to using playlist data as a source of music
similarity information and investigate three approaches to generate anchor and
positive track pairs. We evaluate these approaches by fine-tuning the
pre-trained models for music multi-label classification tasks (genre, mood, and
instrument tagging) and music similarity. We find that creating anchor and
positive track pairs by relying on co-occurrences in playlists provides better
music similarity and competitive classification results compared to choosing
tracks from the same artist as in previous works. Additionally, our best
pre-training approach based on playlists provides superior classification
performance for most datasets.
| no_new_dataset | 0.711469 |
2304.12272 | Young-Suk Lee Dr. | Young-Suk Lee, Ram\'on Fernandez Astudillo, Radu Florian, Tahira
Naseem, Salim Roukos | AMR Parsing with Instruction Fine-tuned Pre-trained Language Models | null | null | null | null | cs.CL cs.AI | http://creativecommons.org/licenses/by-sa/4.0/ | Instruction fine-tuned language models on a collection of instruction
annotated datasets (FLAN) have shown highly effective to improve model
performance and generalization to unseen tasks. However, a majority of standard
parsing tasks including abstract meaning representation (AMR), universal
dependency (UD), semantic role labeling (SRL) has been excluded from the FLAN
collections for both model training and evaluations. In this paper, we take one
of such instruction fine-tuned pre-trained language models, i.e. FLAN-T5, and
fine-tune them for AMR parsing. Our extensive experiments on various AMR
parsing tasks including AMR2.0, AMR3.0 and BioAMR indicate that FLAN-T5
fine-tuned models out-perform previous state-of-the-art models across all
tasks. In addition, full fine-tuning followed by the parameter efficient
fine-tuning, LoRA, further improves the model performances, setting new
state-of-the-arts in Smatch on AMR2.0 (86.4), AMR3.0 (84.9) and BioAMR (82.3).
| [
{
"version": "v1",
"created": "Mon, 24 Apr 2023 17:12:17 GMT"
}
] | 2023-04-25T00:00:00 | [
[
"Lee",
"Young-Suk",
""
],
[
"Astudillo",
"Ramón Fernandez",
""
],
[
"Florian",
"Radu",
""
],
[
"Naseem",
"Tahira",
""
],
[
"Roukos",
"Salim",
""
]
] | TITLE: AMR Parsing with Instruction Fine-tuned Pre-trained Language Models
ABSTRACT: Instruction fine-tuned language models on a collection of instruction
annotated datasets (FLAN) have shown highly effective to improve model
performance and generalization to unseen tasks. However, a majority of standard
parsing tasks including abstract meaning representation (AMR), universal
dependency (UD), semantic role labeling (SRL) has been excluded from the FLAN
collections for both model training and evaluations. In this paper, we take one
of such instruction fine-tuned pre-trained language models, i.e. FLAN-T5, and
fine-tune them for AMR parsing. Our extensive experiments on various AMR
parsing tasks including AMR2.0, AMR3.0 and BioAMR indicate that FLAN-T5
fine-tuned models out-perform previous state-of-the-art models across all
tasks. In addition, full fine-tuning followed by the parameter efficient
fine-tuning, LoRA, further improves the model performances, setting new
state-of-the-arts in Smatch on AMR2.0 (86.4), AMR3.0 (84.9) and BioAMR (82.3).
| no_new_dataset | 0.709416 |
2304.12281 | Jia-Wei Liu | Jia-Wei Liu, Yan-Pei Cao, Tianyuan Yang, Eric Zhongcong Xu, Jussi
Keppo, Ying Shan, Xiaohu Qie, Mike Zheng Shou | HOSNeRF: Dynamic Human-Object-Scene Neural Radiance Fields from a Single
Video | Project page: https://showlab.github.io/HOSNeRF | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We introduce HOSNeRF, a novel 360{\deg} free-viewpoint rendering method that
reconstructs neural radiance fields for dynamic human-object-scene from a
single monocular in-the-wild video. Our method enables pausing the video at any
frame and rendering all scene details (dynamic humans, objects, and
backgrounds) from arbitrary viewpoints. The first challenge in this task is the
complex object motions in human-object interactions, which we tackle by
introducing the new object bones into the conventional human skeleton hierarchy
to effectively estimate large object deformations in our dynamic human-object
model. The second challenge is that humans interact with different objects at
different times, for which we introduce two new learnable object state
embeddings that can be used as conditions for learning our human-object
representation and scene representation, respectively. Extensive experiments
show that HOSNeRF significantly outperforms SOTA approaches on two challenging
datasets by a large margin of 40% ~ 50% in terms of LPIPS. The code, data, and
compelling examples of 360{\deg} free-viewpoint renderings from single videos
will be released in https://showlab.github.io/HOSNeRF.
| [
{
"version": "v1",
"created": "Mon, 24 Apr 2023 17:21:49 GMT"
}
] | 2023-04-25T00:00:00 | [
[
"Liu",
"Jia-Wei",
""
],
[
"Cao",
"Yan-Pei",
""
],
[
"Yang",
"Tianyuan",
""
],
[
"Xu",
"Eric Zhongcong",
""
],
[
"Keppo",
"Jussi",
""
],
[
"Shan",
"Ying",
""
],
[
"Qie",
"Xiaohu",
""
],
[
"Shou",
"Mike Zheng",
""
]
] | TITLE: HOSNeRF: Dynamic Human-Object-Scene Neural Radiance Fields from a Single
Video
ABSTRACT: We introduce HOSNeRF, a novel 360{\deg} free-viewpoint rendering method that
reconstructs neural radiance fields for dynamic human-object-scene from a
single monocular in-the-wild video. Our method enables pausing the video at any
frame and rendering all scene details (dynamic humans, objects, and
backgrounds) from arbitrary viewpoints. The first challenge in this task is the
complex object motions in human-object interactions, which we tackle by
introducing the new object bones into the conventional human skeleton hierarchy
to effectively estimate large object deformations in our dynamic human-object
model. The second challenge is that humans interact with different objects at
different times, for which we introduce two new learnable object state
embeddings that can be used as conditions for learning our human-object
representation and scene representation, respectively. Extensive experiments
show that HOSNeRF significantly outperforms SOTA approaches on two challenging
datasets by a large margin of 40% ~ 50% in terms of LPIPS. The code, data, and
compelling examples of 360{\deg} free-viewpoint renderings from single videos
will be released in https://showlab.github.io/HOSNeRF.
| no_new_dataset | 0.711481 |
2304.12301 | Kun He | Takehiko Ohkawa, Kun He, Fadime Sener, Tomas Hodan, Luan Tran, Cem
Keskin | AssemblyHands: Towards Egocentric Activity Understanding via 3D Hand
Pose Estimation | CVPR 2023. Project page: https://assemblyhands.github.io/ | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | We present AssemblyHands, a large-scale benchmark dataset with accurate 3D
hand pose annotations, to facilitate the study of egocentric activities with
challenging hand-object interactions. The dataset includes synchronized
egocentric and exocentric images sampled from the recent Assembly101 dataset,
in which participants assemble and disassemble take-apart toys. To obtain
high-quality 3D hand pose annotations for the egocentric images, we develop an
efficient pipeline, where we use an initial set of manual annotations to train
a model to automatically annotate a much larger dataset. Our annotation model
uses multi-view feature fusion and an iterative refinement scheme, and achieves
an average keypoint error of 4.20 mm, which is 85% lower than the error of the
original annotations in Assembly101. AssemblyHands provides 3.0M annotated
images, including 490K egocentric images, making it the largest existing
benchmark dataset for egocentric 3D hand pose estimation. Using this data, we
develop a strong single-view baseline of 3D hand pose estimation from
egocentric images. Furthermore, we design a novel action classification task to
evaluate predicted 3D hand poses. Our study shows that having higher-quality
hand poses directly improves the ability to recognize actions.
| [
{
"version": "v1",
"created": "Mon, 24 Apr 2023 17:52:57 GMT"
}
] | 2023-04-25T00:00:00 | [
[
"Ohkawa",
"Takehiko",
""
],
[
"He",
"Kun",
""
],
[
"Sener",
"Fadime",
""
],
[
"Hodan",
"Tomas",
""
],
[
"Tran",
"Luan",
""
],
[
"Keskin",
"Cem",
""
]
] | TITLE: AssemblyHands: Towards Egocentric Activity Understanding via 3D Hand
Pose Estimation
ABSTRACT: We present AssemblyHands, a large-scale benchmark dataset with accurate 3D
hand pose annotations, to facilitate the study of egocentric activities with
challenging hand-object interactions. The dataset includes synchronized
egocentric and exocentric images sampled from the recent Assembly101 dataset,
in which participants assemble and disassemble take-apart toys. To obtain
high-quality 3D hand pose annotations for the egocentric images, we develop an
efficient pipeline, where we use an initial set of manual annotations to train
a model to automatically annotate a much larger dataset. Our annotation model
uses multi-view feature fusion and an iterative refinement scheme, and achieves
an average keypoint error of 4.20 mm, which is 85% lower than the error of the
original annotations in Assembly101. AssemblyHands provides 3.0M annotated
images, including 490K egocentric images, making it the largest existing
benchmark dataset for egocentric 3D hand pose estimation. Using this data, we
develop a strong single-view baseline of 3D hand pose estimation from
egocentric images. Furthermore, we design a novel action classification task to
evaluate predicted 3D hand poses. Our study shows that having higher-quality
hand poses directly improves the ability to recognize actions.
| new_dataset | 0.716188 |
1912.08189 | Nicholas Perello | Przemyslaw A. Grabowicz, Nicholas Perello, Kenta Takatsu | Learning from Discriminatory Training Data | 16 pages, 14 figures, 1 table | null | null | null | cs.LG cs.CY physics.soc-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Supervised learning systems are trained using historical data and, if the
data was tainted by discrimination, they may unintentionally learn to
discriminate against protected groups. We propose that fair learning methods,
despite training on potentially discriminatory datasets, shall perform well on
fair test datasets. Such dataset shifts crystallize application scenarios for
specific fair learning methods. For instance, the removal of direct
discrimination can be represented as a particular dataset shift problem. For
this scenario, we propose a learning method that provably minimizes model error
on fair datasets, while blindly training on datasets poisoned with direct
additive discrimination. The method is compatible with existing legal systems
and provides a solution to the widely discussed issue of protected groups'
intersectionality by striking a balance between the protected groups.
Technically, the method applies probabilistic interventions, has causal and
counterfactual formulations, and is computationally lightweight - it can be
used with any supervised learning model to prevent discrimination via proxies
while maximizing model accuracy for business necessity.
| [
{
"version": "v1",
"created": "Tue, 17 Dec 2019 18:53:23 GMT"
},
{
"version": "v2",
"created": "Mon, 24 Feb 2020 18:50:51 GMT"
},
{
"version": "v3",
"created": "Tue, 23 Feb 2021 02:40:29 GMT"
},
{
"version": "v4",
"created": "Fri, 21 Apr 2023 02:31:12 GMT"
}
] | 2023-04-24T00:00:00 | [
[
"Grabowicz",
"Przemyslaw A.",
""
],
[
"Perello",
"Nicholas",
""
],
[
"Takatsu",
"Kenta",
""
]
] | TITLE: Learning from Discriminatory Training Data
ABSTRACT: Supervised learning systems are trained using historical data and, if the
data was tainted by discrimination, they may unintentionally learn to
discriminate against protected groups. We propose that fair learning methods,
despite training on potentially discriminatory datasets, shall perform well on
fair test datasets. Such dataset shifts crystallize application scenarios for
specific fair learning methods. For instance, the removal of direct
discrimination can be represented as a particular dataset shift problem. For
this scenario, we propose a learning method that provably minimizes model error
on fair datasets, while blindly training on datasets poisoned with direct
additive discrimination. The method is compatible with existing legal systems
and provides a solution to the widely discussed issue of protected groups'
intersectionality by striking a balance between the protected groups.
Technically, the method applies probabilistic interventions, has causal and
counterfactual formulations, and is computationally lightweight - it can be
used with any supervised learning model to prevent discrimination via proxies
while maximizing model accuracy for business necessity.
| no_new_dataset | 0.70883 |
2010.10682 | Hojjat Aghakhani | Hojjat Aghakhani, Lea Sch\"onherr, Thorsten Eisenhofer, Dorothea
Kolossa, Thorsten Holz, Christopher Kruegel, and Giovanni Vigna | VenoMave: Targeted Poisoning Against Speech Recognition | null | null | null | null | cs.SD cs.CR cs.LG eess.AS | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Despite remarkable improvements, automatic speech recognition is susceptible
to adversarial perturbations. Compared to standard machine learning
architectures, these attacks are significantly more challenging, especially
since the inputs to a speech recognition system are time series that contain
both acoustic and linguistic properties of speech. Extracting all
recognition-relevant information requires more complex pipelines and an
ensemble of specialized components. Consequently, an attacker needs to consider
the entire pipeline. In this paper, we present VENOMAVE, the first
training-time poisoning attack against speech recognition. Similar to the
predominantly studied evasion attacks, we pursue the same goal: leading the
system to an incorrect and attacker-chosen transcription of a target audio
waveform. In contrast to evasion attacks, however, we assume that the attacker
can only manipulate a small part of the training data without altering the
target audio waveform at runtime. We evaluate our attack on two datasets:
TIDIGITS and Speech Commands. When poisoning less than 0.17% of the dataset,
VENOMAVE achieves attack success rates of more than 80.0%, without access to
the victim's network architecture or hyperparameters. In a more realistic
scenario, when the target audio waveform is played over the air in different
rooms, VENOMAVE maintains a success rate of up to 73.3%. Finally, VENOMAVE
achieves an attack transferability rate of 36.4% between two different model
architectures.
| [
{
"version": "v1",
"created": "Wed, 21 Oct 2020 00:30:08 GMT"
},
{
"version": "v2",
"created": "Mon, 25 Oct 2021 17:28:34 GMT"
},
{
"version": "v3",
"created": "Thu, 20 Apr 2023 21:21:04 GMT"
}
] | 2023-04-24T00:00:00 | [
[
"Aghakhani",
"Hojjat",
""
],
[
"Schönherr",
"Lea",
""
],
[
"Eisenhofer",
"Thorsten",
""
],
[
"Kolossa",
"Dorothea",
""
],
[
"Holz",
"Thorsten",
""
],
[
"Kruegel",
"Christopher",
""
],
[
"Vigna",
"Giovanni",
""
]
] | TITLE: VenoMave: Targeted Poisoning Against Speech Recognition
ABSTRACT: Despite remarkable improvements, automatic speech recognition is susceptible
to adversarial perturbations. Compared to standard machine learning
architectures, these attacks are significantly more challenging, especially
since the inputs to a speech recognition system are time series that contain
both acoustic and linguistic properties of speech. Extracting all
recognition-relevant information requires more complex pipelines and an
ensemble of specialized components. Consequently, an attacker needs to consider
the entire pipeline. In this paper, we present VENOMAVE, the first
training-time poisoning attack against speech recognition. Similar to the
predominantly studied evasion attacks, we pursue the same goal: leading the
system to an incorrect and attacker-chosen transcription of a target audio
waveform. In contrast to evasion attacks, however, we assume that the attacker
can only manipulate a small part of the training data without altering the
target audio waveform at runtime. We evaluate our attack on two datasets:
TIDIGITS and Speech Commands. When poisoning less than 0.17% of the dataset,
VENOMAVE achieves attack success rates of more than 80.0%, without access to
the victim's network architecture or hyperparameters. In a more realistic
scenario, when the target audio waveform is played over the air in different
rooms, VENOMAVE maintains a success rate of up to 73.3%. Finally, VENOMAVE
achieves an attack transferability rate of 36.4% between two different model
architectures.
| no_new_dataset | 0.709831 |
2012.08987 | Hanlei Zhang | Hanlei Zhang, Hua Xu, Ting-En Lin, Rui Lyu | Discovering New Intents with Deep Aligned Clustering | Accepted by AAAI 2021 (Main Track, Long Paper) | null | 10.1609/aaai.v35i16.17689 | null | cs.CL cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Discovering new intents is a crucial task in dialogue systems. Most existing
methods are limited in transferring the prior knowledge from known intents to
new intents. They also have difficulties in providing high-quality supervised
signals to learn clustering-friendly features for grouping unlabeled intents.
In this work, we propose an effective method, Deep Aligned Clustering, to
discover new intents with the aid of the limited known intent data. Firstly, we
leverage a few labeled known intent samples as prior knowledge to pre-train the
model. Then, we perform k-means to produce cluster assignments as
pseudo-labels. Moreover, we propose an alignment strategy to tackle the label
inconsistency problem during clustering assignments. Finally, we learn the
intent representations under the supervision of the aligned pseudo-labels. With
an unknown number of new intents, we predict the number of intent categories by
eliminating low-confidence intent-wise clusters. Extensive experiments on two
benchmark datasets show that our method is more robust and achieves substantial
improvements over the state-of-the-art methods. The codes are released at
https://github.com/thuiar/DeepAligned-Clustering.
| [
{
"version": "v1",
"created": "Wed, 16 Dec 2020 14:32:06 GMT"
},
{
"version": "v2",
"created": "Fri, 18 Dec 2020 09:45:17 GMT"
},
{
"version": "v3",
"created": "Mon, 21 Dec 2020 01:19:56 GMT"
},
{
"version": "v4",
"created": "Tue, 22 Dec 2020 02:20:28 GMT"
},
{
"version": "v5",
"created": "Tue, 9 Feb 2021 15:27:51 GMT"
},
{
"version": "v6",
"created": "Fri, 19 Mar 2021 03:03:33 GMT"
},
{
"version": "v7",
"created": "Mon, 22 Mar 2021 02:35:17 GMT"
}
] | 2023-04-24T00:00:00 | [
[
"Zhang",
"Hanlei",
""
],
[
"Xu",
"Hua",
""
],
[
"Lin",
"Ting-En",
""
],
[
"Lyu",
"Rui",
""
]
] | TITLE: Discovering New Intents with Deep Aligned Clustering
ABSTRACT: Discovering new intents is a crucial task in dialogue systems. Most existing
methods are limited in transferring the prior knowledge from known intents to
new intents. They also have difficulties in providing high-quality supervised
signals to learn clustering-friendly features for grouping unlabeled intents.
In this work, we propose an effective method, Deep Aligned Clustering, to
discover new intents with the aid of the limited known intent data. Firstly, we
leverage a few labeled known intent samples as prior knowledge to pre-train the
model. Then, we perform k-means to produce cluster assignments as
pseudo-labels. Moreover, we propose an alignment strategy to tackle the label
inconsistency problem during clustering assignments. Finally, we learn the
intent representations under the supervision of the aligned pseudo-labels. With
an unknown number of new intents, we predict the number of intent categories by
eliminating low-confidence intent-wise clusters. Extensive experiments on two
benchmark datasets show that our method is more robust and achieves substantial
improvements over the state-of-the-art methods. The codes are released at
https://github.com/thuiar/DeepAligned-Clustering.
| no_new_dataset | 0.710415 |
2012.10209 | Hanlei Zhang | Hanlei Zhang, Hua Xu, Ting-En Lin | Deep Open Intent Classification with Adaptive Decision Boundary | Accepted by AAAI 2021 (Main Track, Long Paper) | null | 10.1609/aaai.v35i16.17690 | null | cs.CL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Open intent classification is a challenging task in dialogue systems. On the
one hand, it should ensure the quality of known intent identification. On the
other hand, it needs to detect the open (unknown) intent without prior
knowledge. Current models are limited in finding the appropriate decision
boundary to balance the performances of both known intents and the open intent.
In this paper, we propose a post-processing method to learn the adaptive
decision boundary (ADB) for open intent classification. We first utilize the
labeled known intent samples to pre-train the model. Then, we automatically
learn the adaptive spherical decision boundary for each known class with the
aid of well-trained features. Specifically, we propose a new loss function to
balance both the empirical risk and the open space risk. Our method does not
need open intent samples and is free from modifying the model architecture.
Moreover, our approach is surprisingly insensitive with less labeled data and
fewer known intents. Extensive experiments on three benchmark datasets show
that our method yields significant improvements compared with the
state-of-the-art methods. The codes are released at
https://github.com/thuiar/Adaptive-Decision-Boundary.
| [
{
"version": "v1",
"created": "Fri, 18 Dec 2020 13:05:11 GMT"
},
{
"version": "v2",
"created": "Mon, 21 Dec 2020 07:53:55 GMT"
},
{
"version": "v3",
"created": "Wed, 23 Dec 2020 09:19:11 GMT"
},
{
"version": "v4",
"created": "Thu, 11 Feb 2021 02:34:12 GMT"
},
{
"version": "v5",
"created": "Thu, 1 Apr 2021 13:27:49 GMT"
}
] | 2023-04-24T00:00:00 | [
[
"Zhang",
"Hanlei",
""
],
[
"Xu",
"Hua",
""
],
[
"Lin",
"Ting-En",
""
]
] | TITLE: Deep Open Intent Classification with Adaptive Decision Boundary
ABSTRACT: Open intent classification is a challenging task in dialogue systems. On the
one hand, it should ensure the quality of known intent identification. On the
other hand, it needs to detect the open (unknown) intent without prior
knowledge. Current models are limited in finding the appropriate decision
boundary to balance the performances of both known intents and the open intent.
In this paper, we propose a post-processing method to learn the adaptive
decision boundary (ADB) for open intent classification. We first utilize the
labeled known intent samples to pre-train the model. Then, we automatically
learn the adaptive spherical decision boundary for each known class with the
aid of well-trained features. Specifically, we propose a new loss function to
balance both the empirical risk and the open space risk. Our method does not
need open intent samples and is free from modifying the model architecture.
Moreover, our approach is surprisingly insensitive with less labeled data and
fewer known intents. Extensive experiments on three benchmark datasets show
that our method yields significant improvements compared with the
state-of-the-art methods. The codes are released at
https://github.com/thuiar/Adaptive-Decision-Boundary.
| no_new_dataset | 0.710616 |
2104.02260 | Bin Li | Bin Li, Panpan Zhang, Jinye Peng, Hong Fu | Non-contact PPG Signal and Heart Rate Estimation with Multi-hierarchical
Convolutional Network | 34 pages,8 figures | null | 10.1016/j.patcog.2023.109421 | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | Heartbeat rhythm and heart rate (HR) are important physiological parameters
of the human body. This study presents an efficient multi-hierarchical
spatio-temporal convolutional network that can quickly estimate remote
physiological (rPPG) signal and HR from face video clips. First, the facial
color distribution characteristics are extracted using a low-level face feature
generation (LFFG) module. Then, the three-dimensional (3D) spatio-temporal
stack convolution module (STSC) and multi-hierarchical feature fusion module
(MHFF) are used to strengthen the spatio-temporal correlation of multi-channel
features. In the MHFF, sparse optical flow is used to capture the tiny motion
information of faces between frames and generate a self-adaptive region of
interest (ROI) skin mask. Finally, the signal prediction module (SP) is used to
extract the estimated rPPG signal. The heart rate estimation results show that
the proposed network overperforms the state-of-the-art methods on three
datasets, 1) UBFC-RPPG, 2) COHFACE, 3) our dataset, with the mean absolute
error (MAE) of 2.15, 5.57, 1.75 beats per minute (bpm) respectively.
| [
{
"version": "v1",
"created": "Tue, 6 Apr 2021 03:04:27 GMT"
},
{
"version": "v2",
"created": "Fri, 21 Apr 2023 15:03:09 GMT"
}
] | 2023-04-24T00:00:00 | [
[
"Li",
"Bin",
""
],
[
"Zhang",
"Panpan",
""
],
[
"Peng",
"Jinye",
""
],
[
"Fu",
"Hong",
""
]
] | TITLE: Non-contact PPG Signal and Heart Rate Estimation with Multi-hierarchical
Convolutional Network
ABSTRACT: Heartbeat rhythm and heart rate (HR) are important physiological parameters
of the human body. This study presents an efficient multi-hierarchical
spatio-temporal convolutional network that can quickly estimate remote
physiological (rPPG) signal and HR from face video clips. First, the facial
color distribution characteristics are extracted using a low-level face feature
generation (LFFG) module. Then, the three-dimensional (3D) spatio-temporal
stack convolution module (STSC) and multi-hierarchical feature fusion module
(MHFF) are used to strengthen the spatio-temporal correlation of multi-channel
features. In the MHFF, sparse optical flow is used to capture the tiny motion
information of faces between frames and generate a self-adaptive region of
interest (ROI) skin mask. Finally, the signal prediction module (SP) is used to
extract the estimated rPPG signal. The heart rate estimation results show that
the proposed network overperforms the state-of-the-art methods on three
datasets, 1) UBFC-RPPG, 2) COHFACE, 3) our dataset, with the mean absolute
error (MAE) of 2.15, 5.57, 1.75 beats per minute (bpm) respectively.
| no_new_dataset | 0.700844 |
2106.02023 | Patrick James Roddy | Patrick J. Roddy, Jason D. McEwen | Slepian Scale-Discretised Wavelets on the Sphere | 12 pages, 14 figures | IEEE Transactions on Signal Processing, vol. 70, pp. 6142-6153,
2022 | 10.1109/TSP.2022.3233309 | null | cs.IT astro-ph.IM cs.NA math.IT math.NA | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | This work presents the construction of a novel spherical wavelet basis
designed for incomplete spherical datasets, i.e. datasets which are missing in
a particular region of the sphere. The eigenfunctions of the Slepian
spatial-spectral concentration problem (the Slepian functions) are a set of
orthogonal basis functions which are more concentrated within a defined region.
Slepian functions allow one to compute a convolution on the incomplete sphere
by leveraging the recently proposed sifting convolution and extending it to any
set of basis functions. Through a tiling of the Slepian harmonic line, one may
construct scale-discretised wavelets. An illustration is presented based on an
example region on the sphere defined by the topographic map of the Earth. The
Slepian wavelets and corresponding wavelet coefficients are constructed from
this region and are used in a straightforward denoising example.
| [
{
"version": "v1",
"created": "Thu, 3 Jun 2021 17:44:09 GMT"
},
{
"version": "v2",
"created": "Fri, 23 Dec 2022 16:38:51 GMT"
}
] | 2023-04-24T00:00:00 | [
[
"Roddy",
"Patrick J.",
""
],
[
"McEwen",
"Jason D.",
""
]
] | TITLE: Slepian Scale-Discretised Wavelets on the Sphere
ABSTRACT: This work presents the construction of a novel spherical wavelet basis
designed for incomplete spherical datasets, i.e. datasets which are missing in
a particular region of the sphere. The eigenfunctions of the Slepian
spatial-spectral concentration problem (the Slepian functions) are a set of
orthogonal basis functions which are more concentrated within a defined region.
Slepian functions allow one to compute a convolution on the incomplete sphere
by leveraging the recently proposed sifting convolution and extending it to any
set of basis functions. Through a tiling of the Slepian harmonic line, one may
construct scale-discretised wavelets. An illustration is presented based on an
example region on the sphere defined by the topographic map of the Earth. The
Slepian wavelets and corresponding wavelet coefficients are constructed from
this region and are used in a straightforward denoising example.
| no_new_dataset | 0.710427 |
2201.10001 | Ye Gao | Ye Gao, Brian Baucom, Karen Rose, Kristina Gordon, Hongning Wang, John
Stankovic | E-ADDA: Unsupervised Adversarial Domain Adaptation Enhanced by a New
Mahalanobis Distance Loss for Smart Computing | null | null | null | null | cs.LG cs.AI | http://creativecommons.org/licenses/by/4.0/ | In smart computing, the labels of training samples for a specific task are
not always abundant. However, the labels of samples in a relevant but different
dataset are available. As a result, researchers have relied on unsupervised
domain adaptation to leverage the labels in a dataset (the source domain) to
perform better classification in a different, unlabeled dataset (target
domain). Existing non-generative adversarial solutions for UDA aim at achieving
domain confusion through adversarial training. The ideal scenario is that
perfect domain confusion is achieved, but this is not guaranteed to be true. To
further enforce domain confusion on top of the adversarial training, we propose
a novel UDA algorithm, \textit{E-ADDA}, which uses both a novel variation of
the Mahalanobis distance loss and an out-of-distribution detection subroutine.
The Mahalanobis distance loss minimizes the distribution-wise distance between
the encoded target samples and the distribution of the source domain, thus
enforcing additional domain confusion on top of adversarial training. Then, the
OOD subroutine further eliminates samples on which the domain confusion is
unsuccessful. We have performed extensive and comprehensive evaluations of
E-ADDA in the acoustic and computer vision modalities. In the acoustic
modality, E-ADDA outperforms several state-of-the-art UDA algorithms by up to
29.8%, measured in the f1 score. In the computer vision modality, the
evaluation results suggest that we achieve new state-of-the-art performance on
popular UDA benchmarks such as Office-31 and Office-Home, outperforming the
second best-performing algorithms by up to 17.9%.
| [
{
"version": "v1",
"created": "Mon, 24 Jan 2022 23:20:55 GMT"
},
{
"version": "v2",
"created": "Wed, 16 Feb 2022 10:51:08 GMT"
},
{
"version": "v3",
"created": "Mon, 15 Aug 2022 04:21:10 GMT"
},
{
"version": "v4",
"created": "Wed, 9 Nov 2022 22:36:31 GMT"
},
{
"version": "v5",
"created": "Fri, 21 Apr 2023 15:53:46 GMT"
}
] | 2023-04-24T00:00:00 | [
[
"Gao",
"Ye",
""
],
[
"Baucom",
"Brian",
""
],
[
"Rose",
"Karen",
""
],
[
"Gordon",
"Kristina",
""
],
[
"Wang",
"Hongning",
""
],
[
"Stankovic",
"John",
""
]
] | TITLE: E-ADDA: Unsupervised Adversarial Domain Adaptation Enhanced by a New
Mahalanobis Distance Loss for Smart Computing
ABSTRACT: In smart computing, the labels of training samples for a specific task are
not always abundant. However, the labels of samples in a relevant but different
dataset are available. As a result, researchers have relied on unsupervised
domain adaptation to leverage the labels in a dataset (the source domain) to
perform better classification in a different, unlabeled dataset (target
domain). Existing non-generative adversarial solutions for UDA aim at achieving
domain confusion through adversarial training. The ideal scenario is that
perfect domain confusion is achieved, but this is not guaranteed to be true. To
further enforce domain confusion on top of the adversarial training, we propose
a novel UDA algorithm, \textit{E-ADDA}, which uses both a novel variation of
the Mahalanobis distance loss and an out-of-distribution detection subroutine.
The Mahalanobis distance loss minimizes the distribution-wise distance between
the encoded target samples and the distribution of the source domain, thus
enforcing additional domain confusion on top of adversarial training. Then, the
OOD subroutine further eliminates samples on which the domain confusion is
unsuccessful. We have performed extensive and comprehensive evaluations of
E-ADDA in the acoustic and computer vision modalities. In the acoustic
modality, E-ADDA outperforms several state-of-the-art UDA algorithms by up to
29.8%, measured in the f1 score. In the computer vision modality, the
evaluation results suggest that we achieve new state-of-the-art performance on
popular UDA benchmarks such as Office-31 and Office-Home, outperforming the
second best-performing algorithms by up to 17.9%.
| no_new_dataset | 0.7114 |
2202.03628 | Zihao Xu | Zihao Xu, Hao He, Guang-He Lee, Yuyang Wang, Hao Wang | Graph-Relational Domain Adaptation | Accepted by ICLR 2022 | null | null | null | cs.LG cs.AI cs.CV cs.SI stat.ML | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Existing domain adaptation methods tend to treat every domain equally and
align them all perfectly. Such uniform alignment ignores topological structures
among different domains; therefore it may be beneficial for nearby domains, but
not necessarily for distant domains. In this work, we relax such uniform
alignment by using a domain graph to encode domain adjacency, e.g., a graph of
states in the US with each state as a domain and each edge indicating
adjacency, thereby allowing domains to align flexibly based on the graph
structure. We generalize the existing adversarial learning framework with a
novel graph discriminator using encoding-conditioned graph embeddings.
Theoretical analysis shows that at equilibrium, our method recovers classic
domain adaptation when the graph is a clique, and achieves non-trivial
alignment for other types of graphs. Empirical results show that our approach
successfully generalizes uniform alignment, naturally incorporates domain
information represented by graphs, and improves upon existing domain adaptation
methods on both synthetic and real-world datasets. Code will soon be available
at https://github.com/Wang-ML-Lab/GRDA.
| [
{
"version": "v1",
"created": "Tue, 8 Feb 2022 03:53:32 GMT"
},
{
"version": "v2",
"created": "Fri, 21 Apr 2023 02:40:15 GMT"
}
] | 2023-04-24T00:00:00 | [
[
"Xu",
"Zihao",
""
],
[
"He",
"Hao",
""
],
[
"Lee",
"Guang-He",
""
],
[
"Wang",
"Yuyang",
""
],
[
"Wang",
"Hao",
""
]
] | TITLE: Graph-Relational Domain Adaptation
ABSTRACT: Existing domain adaptation methods tend to treat every domain equally and
align them all perfectly. Such uniform alignment ignores topological structures
among different domains; therefore it may be beneficial for nearby domains, but
not necessarily for distant domains. In this work, we relax such uniform
alignment by using a domain graph to encode domain adjacency, e.g., a graph of
states in the US with each state as a domain and each edge indicating
adjacency, thereby allowing domains to align flexibly based on the graph
structure. We generalize the existing adversarial learning framework with a
novel graph discriminator using encoding-conditioned graph embeddings.
Theoretical analysis shows that at equilibrium, our method recovers classic
domain adaptation when the graph is a clique, and achieves non-trivial
alignment for other types of graphs. Empirical results show that our approach
successfully generalizes uniform alignment, naturally incorporates domain
information represented by graphs, and improves upon existing domain adaptation
methods on both synthetic and real-world datasets. Code will soon be available
at https://github.com/Wang-ML-Lab/GRDA.
| no_new_dataset | 0.712401 |
2202.06608 | Nico Weber | Nico Weber, Christoph Thiem, and Ulrich Konigorski | Toward Unsupervised Test Scenario Extraction for Automated Driving
Systems from Urban Naturalistic Road Traffic Data | 16 pages, 9 figures | SAE Intl. J CAV 6(3):2023 | 10.4271/12-06-03-0017 | SAE 12-06-03-0017 | cs.SE cs.LG | http://creativecommons.org/licenses/by-sa/4.0/ | Scenario-based testing is a promising approach to solve the challenge of
proving the safe behavior of vehicles equipped with automated driving systems.
Since an infinite number of concrete scenarios can theoretically occur in
real-world road traffic, the extraction of scenarios relevant in terms of the
safety-related behavior of these systems is a key aspect for their successful
verification and validation. Therefore, a method for extracting multimodal
urban traffic scenarios from naturalistic road traffic data in an unsupervised
manner, minimizing the amount of (potentially biased) prior expert knowledge,
is proposed. Rather than an (elaborate) rule-based assignment by extracting
concrete scenarios into predefined functional scenarios, the presented method
deploys an unsupervised machine learning pipeline. The approach allows
exploring the unknown nature of the data and their interpretation as test
scenarios that experts could not have anticipated. The method is evaluated for
naturalistic road traffic data at urban intersections from the inD and the
Silicon Valley Intersections datasets. For this purpose, it is analyzed with
which clustering approach (K-Means, hierarchical clustering, and DBSCAN) the
scenario extraction method performs best (referring to an elaborate rule-based
implementation). Subsequently, using hierarchical clustering the results show
both a jump in overall accuracy of around 20% when moving from 4 to 5 clusters
and a saturation effect starting at 41 clusters with an overall accuracy of
84%. These observations can be a valuable contribution in the context of the
trade-off between the number of functional scenarios (i.e., clustering
accuracy) and testing effort. Possible reasons for the observed accuracy
variations of different clusters, each with a fixed total number of given
clusters, are discussed.
| [
{
"version": "v1",
"created": "Mon, 14 Feb 2022 10:55:14 GMT"
},
{
"version": "v2",
"created": "Fri, 21 Apr 2023 14:15:57 GMT"
}
] | 2023-04-24T00:00:00 | [
[
"Weber",
"Nico",
""
],
[
"Thiem",
"Christoph",
""
],
[
"Konigorski",
"Ulrich",
""
]
] | TITLE: Toward Unsupervised Test Scenario Extraction for Automated Driving
Systems from Urban Naturalistic Road Traffic Data
ABSTRACT: Scenario-based testing is a promising approach to solve the challenge of
proving the safe behavior of vehicles equipped with automated driving systems.
Since an infinite number of concrete scenarios can theoretically occur in
real-world road traffic, the extraction of scenarios relevant in terms of the
safety-related behavior of these systems is a key aspect for their successful
verification and validation. Therefore, a method for extracting multimodal
urban traffic scenarios from naturalistic road traffic data in an unsupervised
manner, minimizing the amount of (potentially biased) prior expert knowledge,
is proposed. Rather than an (elaborate) rule-based assignment by extracting
concrete scenarios into predefined functional scenarios, the presented method
deploys an unsupervised machine learning pipeline. The approach allows
exploring the unknown nature of the data and their interpretation as test
scenarios that experts could not have anticipated. The method is evaluated for
naturalistic road traffic data at urban intersections from the inD and the
Silicon Valley Intersections datasets. For this purpose, it is analyzed with
which clustering approach (K-Means, hierarchical clustering, and DBSCAN) the
scenario extraction method performs best (referring to an elaborate rule-based
implementation). Subsequently, using hierarchical clustering the results show
both a jump in overall accuracy of around 20% when moving from 4 to 5 clusters
and a saturation effect starting at 41 clusters with an overall accuracy of
84%. These observations can be a valuable contribution in the context of the
trade-off between the number of functional scenarios (i.e., clustering
accuracy) and testing effort. Possible reasons for the observed accuracy
variations of different clusters, each with a fixed total number of given
clusters, are discussed.
| no_new_dataset | 0.713825 |
2202.10075 | Constantine Doumanidis | Constantine Doumanidis (1), Prashant Hari Narayan Rajput (2), Michail
Maniatakos (1) ((1) New York University Abu Dhabi, (2) NYU Tandon School of
Engineering) | ICSML: Industrial Control Systems ML Framework for native inference
using IEC 61131-3 code | 12 pages, 8 figures, code available at
https://github.com/momalab/ICSML, to appear in CPSS 2023 workshop (ACM
AsiaCCS'23) | null | null | null | cs.LG cs.CR cs.SY eess.SY | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Industrial Control Systems (ICS) have played a catalytic role in enabling the
4th Industrial Revolution. ICS devices like Programmable Logic Controllers
(PLCs), automate, monitor, and control critical processes in industrial,
energy, and commercial environments. The convergence of traditional Operational
Technology (OT) with Information Technology (IT) has opened a new and unique
threat landscape. This has inspired defense research that focuses heavily on
Machine Learning (ML) based anomaly detection methods that run on external IT
hardware, which means an increase in costs and the further expansion of the
threat landscape. To remove this requirement, we introduce the ICS machine
learning inference framework (ICSML) which enables executing ML model inference
natively on the PLC. ICSML is implemented in IEC 61131-3 code and provides
several optimizations to bypass the limitations imposed by the domain-specific
languages. Therefore, it works on every PLC without the need for vendor
support. ICSML provides a complete set of components for creating full ML
models similarly to established ML frameworks. We run a series of benchmarks
studying memory and performance, and compare our solution to the TFLite
inference framework. At the same time, we develop domain-specific model
optimizations to improve the efficiency of ICSML. To demonstrate the abilities
of ICSML, we evaluate a case study of a real defense for process-aware attacks
targeting a desalination plant.
| [
{
"version": "v1",
"created": "Mon, 21 Feb 2022 09:37:28 GMT"
},
{
"version": "v2",
"created": "Thu, 14 Apr 2022 11:09:37 GMT"
},
{
"version": "v3",
"created": "Fri, 21 Apr 2023 08:25:13 GMT"
}
] | 2023-04-24T00:00:00 | [
[
"Doumanidis",
"Constantine",
""
],
[
"Rajput",
"Prashant Hari Narayan",
""
],
[
"Maniatakos",
"Michail",
""
]
] | TITLE: ICSML: Industrial Control Systems ML Framework for native inference
using IEC 61131-3 code
ABSTRACT: Industrial Control Systems (ICS) have played a catalytic role in enabling the
4th Industrial Revolution. ICS devices like Programmable Logic Controllers
(PLCs), automate, monitor, and control critical processes in industrial,
energy, and commercial environments. The convergence of traditional Operational
Technology (OT) with Information Technology (IT) has opened a new and unique
threat landscape. This has inspired defense research that focuses heavily on
Machine Learning (ML) based anomaly detection methods that run on external IT
hardware, which means an increase in costs and the further expansion of the
threat landscape. To remove this requirement, we introduce the ICS machine
learning inference framework (ICSML) which enables executing ML model inference
natively on the PLC. ICSML is implemented in IEC 61131-3 code and provides
several optimizations to bypass the limitations imposed by the domain-specific
languages. Therefore, it works on every PLC without the need for vendor
support. ICSML provides a complete set of components for creating full ML
models similarly to established ML frameworks. We run a series of benchmarks
studying memory and performance, and compare our solution to the TFLite
inference framework. At the same time, we develop domain-specific model
optimizations to improve the efficiency of ICSML. To demonstrate the abilities
of ICSML, we evaluate a case study of a real defense for process-aware attacks
targeting a desalination plant.
| no_new_dataset | 0.707588 |
2203.05573 | Lei Zhou | Lei Zhou, Huidong Liu, Joseph Bae, Junjun He, Dimitris Samaras,
Prateek Prasanna | Self Pre-training with Masked Autoencoders for Medical Image
Classification and Segmentation | ISBI2023 camera-ready version (no substantial difference from v1);
Code is available at https://github.com/cvlab-stonybrook/SelfMedMAE | null | null | null | eess.IV cs.CV cs.LG | http://creativecommons.org/licenses/by/4.0/ | Masked Autoencoder (MAE) has recently been shown to be effective in
pre-training Vision Transformers (ViT) for natural image analysis. By
reconstructing full images from partially masked inputs, a ViT encoder
aggregates contextual information to infer masked image regions. We believe
that this context aggregation ability is particularly essential to the medical
image domain where each anatomical structure is functionally and mechanically
connected to other structures and regions. Because there is no ImageNet-scale
medical image dataset for pre-training, we investigate a self pre-training
paradigm with MAE for medical image analysis tasks. Our method pre-trains a ViT
on the training set of the target data instead of another dataset. Thus, self
pre-training can benefit more scenarios where pre-training data is hard to
acquire. Our experimental results show that MAE self pre-training markedly
improves diverse medical image tasks including chest X-ray disease
classification, abdominal CT multi-organ segmentation, and MRI brain tumor
segmentation. Code is available at
https://github.com/cvlab-stonybrook/SelfMedMAE
| [
{
"version": "v1",
"created": "Thu, 10 Mar 2022 16:22:38 GMT"
},
{
"version": "v2",
"created": "Fri, 21 Apr 2023 12:40:08 GMT"
}
] | 2023-04-24T00:00:00 | [
[
"Zhou",
"Lei",
""
],
[
"Liu",
"Huidong",
""
],
[
"Bae",
"Joseph",
""
],
[
"He",
"Junjun",
""
],
[
"Samaras",
"Dimitris",
""
],
[
"Prasanna",
"Prateek",
""
]
] | TITLE: Self Pre-training with Masked Autoencoders for Medical Image
Classification and Segmentation
ABSTRACT: Masked Autoencoder (MAE) has recently been shown to be effective in
pre-training Vision Transformers (ViT) for natural image analysis. By
reconstructing full images from partially masked inputs, a ViT encoder
aggregates contextual information to infer masked image regions. We believe
that this context aggregation ability is particularly essential to the medical
image domain where each anatomical structure is functionally and mechanically
connected to other structures and regions. Because there is no ImageNet-scale
medical image dataset for pre-training, we investigate a self pre-training
paradigm with MAE for medical image analysis tasks. Our method pre-trains a ViT
on the training set of the target data instead of another dataset. Thus, self
pre-training can benefit more scenarios where pre-training data is hard to
acquire. Our experimental results show that MAE self pre-training markedly
improves diverse medical image tasks including chest X-ray disease
classification, abdominal CT multi-organ segmentation, and MRI brain tumor
segmentation. Code is available at
https://github.com/cvlab-stonybrook/SelfMedMAE
| no_new_dataset | 0.709019 |
2203.14523 | Yuyuan Liu | Yuyuan Liu, Yu Tian, Chong Wang, Yuanhong Chen, Fengbei Liu, Vasileios
Belagiannis, Gustavo Carneiro | Translation Consistent Semi-supervised Segmentation for 3D Medical
Images | null | null | null | null | cs.CV cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | 3D medical image segmentation methods have been successful, but their
dependence on large amounts of voxel-level annotated data is a disadvantage
that needs to be addressed given the high cost to obtain such annotation.
Semi-supervised learning (SSL) solve this issue by training models with a large
unlabelled and a small labelled dataset. The most successful SSL approaches are
based on consistency learning that minimises the distance between model
responses obtained from perturbed views of the unlabelled data. These
perturbations usually keep the spatial input context between views fairly
consistent, which may cause the model to learn segmentation patterns from the
spatial input contexts instead of the segmented objects. In this paper, we
introduce the Translation Consistent Co-training (TraCoCo) which is a
consistency learning SSL method that perturbs the input data views by varying
their spatial input context, allowing the model to learn segmentation patterns
from visual objects. Furthermore, we propose the replacement of the commonly
used mean squared error (MSE) semi-supervised loss by a new Cross-model
confident Binary Cross entropy (CBC) loss, which improves training convergence
and keeps the robustness to co-training pseudo-labelling mistakes. We also
extend CutMix augmentation to 3D SSL to further improve generalisation. Our
TraCoCo shows state-of-the-art results for the Left Atrium (LA) and Brain Tumor
Segmentation (BRaTS19) datasets with different backbones. Our code is available
at https://github.com/yyliu01/TraCoCo.
| [
{
"version": "v1",
"created": "Mon, 28 Mar 2022 06:31:39 GMT"
},
{
"version": "v2",
"created": "Fri, 21 Apr 2023 07:09:39 GMT"
}
] | 2023-04-24T00:00:00 | [
[
"Liu",
"Yuyuan",
""
],
[
"Tian",
"Yu",
""
],
[
"Wang",
"Chong",
""
],
[
"Chen",
"Yuanhong",
""
],
[
"Liu",
"Fengbei",
""
],
[
"Belagiannis",
"Vasileios",
""
],
[
"Carneiro",
"Gustavo",
""
]
] | TITLE: Translation Consistent Semi-supervised Segmentation for 3D Medical
Images
ABSTRACT: 3D medical image segmentation methods have been successful, but their
dependence on large amounts of voxel-level annotated data is a disadvantage
that needs to be addressed given the high cost to obtain such annotation.
Semi-supervised learning (SSL) solve this issue by training models with a large
unlabelled and a small labelled dataset. The most successful SSL approaches are
based on consistency learning that minimises the distance between model
responses obtained from perturbed views of the unlabelled data. These
perturbations usually keep the spatial input context between views fairly
consistent, which may cause the model to learn segmentation patterns from the
spatial input contexts instead of the segmented objects. In this paper, we
introduce the Translation Consistent Co-training (TraCoCo) which is a
consistency learning SSL method that perturbs the input data views by varying
their spatial input context, allowing the model to learn segmentation patterns
from visual objects. Furthermore, we propose the replacement of the commonly
used mean squared error (MSE) semi-supervised loss by a new Cross-model
confident Binary Cross entropy (CBC) loss, which improves training convergence
and keeps the robustness to co-training pseudo-labelling mistakes. We also
extend CutMix augmentation to 3D SSL to further improve generalisation. Our
TraCoCo shows state-of-the-art results for the Left Atrium (LA) and Brain Tumor
Segmentation (BRaTS19) datasets with different backbones. Our code is available
at https://github.com/yyliu01/TraCoCo.
| no_new_dataset | 0.71145 |
2204.02947 | Nicholas Perello | Przemyslaw Grabowicz, Nicholas Perello, Aarshee Mishra | Marrying Fairness and Explainability in Supervised Learning | 17 pages, 16 figures. Section 4.3 updated | 2022 ACM Conference on Fairness, Accountability, and Transparency
(FAccT '22) | 10.1145/3531146.3533236 | null | cs.LG cs.CY | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Machine learning algorithms that aid human decision-making may inadvertently
discriminate against certain protected groups. We formalize direct
discrimination as a direct causal effect of the protected attributes on the
decisions, while induced discrimination as a change in the causal influence of
non-protected features associated with the protected attributes. The
measurements of marginal direct effect (MDE) and SHapley Additive exPlanations
(SHAP) reveal that state-of-the-art fair learning methods can induce
discrimination via association or reverse discrimination in synthetic and
real-world datasets. To inhibit discrimination in algorithmic systems, we
propose to nullify the influence of the protected attribute on the output of
the system, while preserving the influence of remaining features. We introduce
and study post-processing methods achieving such objectives, finding that they
yield relatively high model accuracy, prevent direct discrimination, and
diminishes various disparity measures, e.g., demographic disparity.
| [
{
"version": "v1",
"created": "Wed, 6 Apr 2022 17:26:58 GMT"
},
{
"version": "v2",
"created": "Sat, 14 May 2022 20:43:20 GMT"
},
{
"version": "v3",
"created": "Fri, 21 Apr 2023 02:49:51 GMT"
}
] | 2023-04-24T00:00:00 | [
[
"Grabowicz",
"Przemyslaw",
""
],
[
"Perello",
"Nicholas",
""
],
[
"Mishra",
"Aarshee",
""
]
] | TITLE: Marrying Fairness and Explainability in Supervised Learning
ABSTRACT: Machine learning algorithms that aid human decision-making may inadvertently
discriminate against certain protected groups. We formalize direct
discrimination as a direct causal effect of the protected attributes on the
decisions, while induced discrimination as a change in the causal influence of
non-protected features associated with the protected attributes. The
measurements of marginal direct effect (MDE) and SHapley Additive exPlanations
(SHAP) reveal that state-of-the-art fair learning methods can induce
discrimination via association or reverse discrimination in synthetic and
real-world datasets. To inhibit discrimination in algorithmic systems, we
propose to nullify the influence of the protected attribute on the output of
the system, while preserving the influence of remaining features. We introduce
and study post-processing methods achieving such objectives, finding that they
yield relatively high model accuracy, prevent direct discrimination, and
diminishes various disparity measures, e.g., demographic disparity.
| no_new_dataset | 0.708843 |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.