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
2209.06015
Yiming Li
Yiming Li, Mingyan Zhu, Xue Yang, Yong Jiang, Tao Wei, Shu-Tao Xia
Black-box Dataset Ownership Verification via Backdoor Watermarking
This paper is accepted by IEEE TIFS. 15 pages. The preliminary short version of this paper was posted on arXiv (arXiv:2010.05821) and presented in a non-archival NeurIPS Workshop (2020)
null
null
null
cs.CR cs.AI cs.CV cs.LG
http://creativecommons.org/licenses/by/4.0/
Deep learning, especially deep neural networks (DNNs), has been widely and successfully adopted in many critical applications for its high effectiveness and efficiency. The rapid development of DNNs has benefited from the existence of some high-quality datasets ($e.g.$, ImageNet), which allow researchers and developers to easily verify the performance of their methods. Currently, almost all existing released datasets require that they can only be adopted for academic or educational purposes rather than commercial purposes without permission. However, there is still no good way to ensure that. In this paper, we formulate the protection of released datasets as verifying whether they are adopted for training a (suspicious) third-party model, where defenders can only query the model while having no information about its parameters and training details. Based on this formulation, we propose to embed external patterns via backdoor watermarking for the ownership verification to protect them. Our method contains two main parts, including dataset watermarking and dataset verification. Specifically, we exploit poison-only backdoor attacks ($e.g.$, BadNets) for dataset watermarking and design a hypothesis-test-guided method for dataset verification. We also provide some theoretical analyses of our methods. Experiments on multiple benchmark datasets of different tasks are conducted, which verify the effectiveness of our method. The code for reproducing main experiments is available at \url{https://github.com/THUYimingLi/DVBW}.
[ { "version": "v1", "created": "Thu, 4 Aug 2022 05:32:20 GMT" }, { "version": "v2", "created": "Fri, 31 Mar 2023 01:11:50 GMT" } ]
2023-04-03T00:00:00
[ [ "Li", "Yiming", "" ], [ "Zhu", "Mingyan", "" ], [ "Yang", "Xue", "" ], [ "Jiang", "Yong", "" ], [ "Wei", "Tao", "" ], [ "Xia", "Shu-Tao", "" ] ]
TITLE: Black-box Dataset Ownership Verification via Backdoor Watermarking ABSTRACT: Deep learning, especially deep neural networks (DNNs), has been widely and successfully adopted in many critical applications for its high effectiveness and efficiency. The rapid development of DNNs has benefited from the existence of some high-quality datasets ($e.g.$, ImageNet), which allow researchers and developers to easily verify the performance of their methods. Currently, almost all existing released datasets require that they can only be adopted for academic or educational purposes rather than commercial purposes without permission. However, there is still no good way to ensure that. In this paper, we formulate the protection of released datasets as verifying whether they are adopted for training a (suspicious) third-party model, where defenders can only query the model while having no information about its parameters and training details. Based on this formulation, we propose to embed external patterns via backdoor watermarking for the ownership verification to protect them. Our method contains two main parts, including dataset watermarking and dataset verification. Specifically, we exploit poison-only backdoor attacks ($e.g.$, BadNets) for dataset watermarking and design a hypothesis-test-guided method for dataset verification. We also provide some theoretical analyses of our methods. Experiments on multiple benchmark datasets of different tasks are conducted, which verify the effectiveness of our method. The code for reproducing main experiments is available at \url{https://github.com/THUYimingLi/DVBW}.
no_new_dataset
0.709265
2210.01633
Michael Cohen
Michael K. Cohen, Samuel Daulton, Michael A. Osborne
Log-Linear-Time Gaussian Processes Using Binary Tree Kernels
NeurIPS 2022; 9 pages + appendices
Adv.Neur.Info.Proc.Sys. 35 (2022) 8118-8129
null
null
cs.LG
http://creativecommons.org/licenses/by/4.0/
Gaussian processes (GPs) produce good probabilistic models of functions, but most GP kernels require $O((n+m)n^2)$ time, where $n$ is the number of data points and $m$ the number of predictive locations. We present a new kernel that allows for Gaussian process regression in $O((n+m)\log(n+m))$ time. Our "binary tree" kernel places all data points on the leaves of a binary tree, with the kernel depending only on the depth of the deepest common ancestor. We can store the resulting kernel matrix in $O(n)$ space in $O(n \log n)$ time, as a sum of sparse rank-one matrices, and approximately invert the kernel matrix in $O(n)$ time. Sparse GP methods also offer linear run time, but they predict less well than higher dimensional kernels. On a classic suite of regression tasks, we compare our kernel against Mat\'ern, sparse, and sparse variational kernels. The binary tree GP assigns the highest likelihood to the test data on a plurality of datasets, usually achieves lower mean squared error than the sparse methods, and often ties or beats the Mat\'ern GP. On large datasets, the binary tree GP is fastest, and much faster than a Mat\'ern GP.
[ { "version": "v1", "created": "Tue, 4 Oct 2022 14:30:06 GMT" } ]
2023-04-03T00:00:00
[ [ "Cohen", "Michael K.", "" ], [ "Daulton", "Samuel", "" ], [ "Osborne", "Michael A.", "" ] ]
TITLE: Log-Linear-Time Gaussian Processes Using Binary Tree Kernels ABSTRACT: Gaussian processes (GPs) produce good probabilistic models of functions, but most GP kernels require $O((n+m)n^2)$ time, where $n$ is the number of data points and $m$ the number of predictive locations. We present a new kernel that allows for Gaussian process regression in $O((n+m)\log(n+m))$ time. Our "binary tree" kernel places all data points on the leaves of a binary tree, with the kernel depending only on the depth of the deepest common ancestor. We can store the resulting kernel matrix in $O(n)$ space in $O(n \log n)$ time, as a sum of sparse rank-one matrices, and approximately invert the kernel matrix in $O(n)$ time. Sparse GP methods also offer linear run time, but they predict less well than higher dimensional kernels. On a classic suite of regression tasks, we compare our kernel against Mat\'ern, sparse, and sparse variational kernels. The binary tree GP assigns the highest likelihood to the test data on a plurality of datasets, usually achieves lower mean squared error than the sparse methods, and often ties or beats the Mat\'ern GP. On large datasets, the binary tree GP is fastest, and much faster than a Mat\'ern GP.
no_new_dataset
0.712807
2210.12113
Pouria Rouzrokh
Pouria Rouzrokh, Bardia Khosravi, Shahriar Faghani, Mana Moassefi, Sanaz Vahdati, Bradley J. Erickson
Multitask Brain Tumor Inpainting with Diffusion Models: A Methodological Report
17 pages, 7 figures
null
null
null
eess.IV cs.CV cs.LG
http://creativecommons.org/licenses/by-nc-sa/4.0/
Despite the ever-increasing interest in applying deep learning (DL) models to medical imaging, the typical scarcity and imbalance of medical datasets can severely impact the performance of DL models. The generation of synthetic data that might be freely shared without compromising patient privacy is a well-known technique for addressing these difficulties. Inpainting algorithms are a subset of DL generative models that can alter one or more regions of an input image while matching its surrounding context and, in certain cases, non-imaging input conditions. Although the majority of inpainting techniques for medical imaging data use generative adversarial networks (GANs), the performance of these algorithms is frequently suboptimal due to their limited output variety, a problem that is already well-known for GANs. Denoising diffusion probabilistic models (DDPMs) are a recently introduced family of generative networks that can generate results of comparable quality to GANs, but with diverse outputs. In this paper, we describe a DDPM to execute multiple inpainting tasks on 2D axial slices of brain MRI with various sequences, and present proof-of-concept examples of its performance in a variety of evaluation scenarios. Our model and a public online interface to try our tool are available at: https://github.com/Mayo-Radiology-Informatics-Lab/MBTI
[ { "version": "v1", "created": "Fri, 21 Oct 2022 17:13:14 GMT" }, { "version": "v2", "created": "Thu, 30 Mar 2023 18:36:27 GMT" } ]
2023-04-03T00:00:00
[ [ "Rouzrokh", "Pouria", "" ], [ "Khosravi", "Bardia", "" ], [ "Faghani", "Shahriar", "" ], [ "Moassefi", "Mana", "" ], [ "Vahdati", "Sanaz", "" ], [ "Erickson", "Bradley J.", "" ] ]
TITLE: Multitask Brain Tumor Inpainting with Diffusion Models: A Methodological Report ABSTRACT: Despite the ever-increasing interest in applying deep learning (DL) models to medical imaging, the typical scarcity and imbalance of medical datasets can severely impact the performance of DL models. The generation of synthetic data that might be freely shared without compromising patient privacy is a well-known technique for addressing these difficulties. Inpainting algorithms are a subset of DL generative models that can alter one or more regions of an input image while matching its surrounding context and, in certain cases, non-imaging input conditions. Although the majority of inpainting techniques for medical imaging data use generative adversarial networks (GANs), the performance of these algorithms is frequently suboptimal due to their limited output variety, a problem that is already well-known for GANs. Denoising diffusion probabilistic models (DDPMs) are a recently introduced family of generative networks that can generate results of comparable quality to GANs, but with diverse outputs. In this paper, we describe a DDPM to execute multiple inpainting tasks on 2D axial slices of brain MRI with various sequences, and present proof-of-concept examples of its performance in a variety of evaluation scenarios. Our model and a public online interface to try our tool are available at: https://github.com/Mayo-Radiology-Informatics-Lab/MBTI
no_new_dataset
0.706222
2211.03726
Carl Doersch
Carl Doersch, Ankush Gupta, Larisa Markeeva, Adri\`a Recasens, Lucas Smaira, Yusuf Aytar, Jo\~ao Carreira, Andrew Zisserman, Yi Yang
TAP-Vid: A Benchmark for Tracking Any Point in a Video
Published in NeurIPS Datasets and Benchmarks track, 2022
null
null
null
cs.CV stat.ML
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
Generic motion understanding from video involves not only tracking objects, but also perceiving how their surfaces deform and move. This information is useful to make inferences about 3D shape, physical properties and object interactions. While the problem of tracking arbitrary physical points on surfaces over longer video clips has received some attention, no dataset or benchmark for evaluation existed, until now. In this paper, we first formalize the problem, naming it tracking any point (TAP). We introduce a companion benchmark, TAP-Vid, which is composed of both real-world videos with accurate human annotations of point tracks, and synthetic videos with perfect ground-truth point tracks. Central to the construction of our benchmark is a novel semi-automatic crowdsourced pipeline which uses optical flow estimates to compensate for easier, short-term motion like camera shake, allowing annotators to focus on harder sections of video. We validate our pipeline on synthetic data and propose a simple end-to-end point tracking model TAP-Net, showing that it outperforms all prior methods on our benchmark when trained on synthetic data.
[ { "version": "v1", "created": "Mon, 7 Nov 2022 17:57:02 GMT" }, { "version": "v2", "created": "Fri, 31 Mar 2023 11:51:40 GMT" } ]
2023-04-03T00:00:00
[ [ "Doersch", "Carl", "" ], [ "Gupta", "Ankush", "" ], [ "Markeeva", "Larisa", "" ], [ "Recasens", "Adrià", "" ], [ "Smaira", "Lucas", "" ], [ "Aytar", "Yusuf", "" ], [ "Carreira", "João", "" ], [ "Zisserman", "Andrew", "" ], [ "Yang", "Yi", "" ] ]
TITLE: TAP-Vid: A Benchmark for Tracking Any Point in a Video ABSTRACT: Generic motion understanding from video involves not only tracking objects, but also perceiving how their surfaces deform and move. This information is useful to make inferences about 3D shape, physical properties and object interactions. While the problem of tracking arbitrary physical points on surfaces over longer video clips has received some attention, no dataset or benchmark for evaluation existed, until now. In this paper, we first formalize the problem, naming it tracking any point (TAP). We introduce a companion benchmark, TAP-Vid, which is composed of both real-world videos with accurate human annotations of point tracks, and synthetic videos with perfect ground-truth point tracks. Central to the construction of our benchmark is a novel semi-automatic crowdsourced pipeline which uses optical flow estimates to compensate for easier, short-term motion like camera shake, allowing annotators to focus on harder sections of video. We validate our pipeline on synthetic data and propose a simple end-to-end point tracking model TAP-Net, showing that it outperforms all prior methods on our benchmark when trained on synthetic data.
no_new_dataset
0.573201
2211.07021
Eddie Bkheet
Eddie Bkheet, Anne-Lise D'Angelo, Adam Goldbraikh, Shlomi Laufer
Using Hand Pose Estimation To Automate Open Surgery Training Feedback
Accepted to IPCAI 2023, 12 pages, 5 figures
null
null
null
cs.CV
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
Purpose: This research aims to facilitate the use of state-of-the-art computer vision algorithms for the automated training of surgeons and the analysis of surgical footage. By estimating 2D hand poses, we model the movement of the practitioner's hands, and their interaction with surgical instruments, to study their potential benefit for surgical training. Methods: We leverage pre-trained models on a publicly-available hands dataset to create our own in-house dataset of 100 open surgery simulation videos with 2D hand poses. We also assess the ability of pose estimations to segment surgical videos into gestures and tool-usage segments and compare them to kinematic sensors and I3D features. Furthermore, we introduce 6 novel surgical dexterity proxies stemming from domain experts' training advice, all of which our framework can automatically detect given raw video footage. Results: State-of-the-art gesture segmentation accuracy of 88.35\% on the Open Surgery Simulation dataset is achieved with the fusion of 2D poses and I3D features from multiple angles. The introduced surgical skill proxies presented significant differences for novices compared to experts and produced actionable feedback for improvement. Conclusion: This research demonstrates the benefit of pose estimations for open surgery by analyzing their effectiveness in gesture segmentation and skill assessment. Gesture segmentation using pose estimations achieved comparable results to physical sensors while being remote and markerless. Surgical dexterity proxies that rely on pose estimation proved they can be used to work towards automated training feedback. We hope our findings encourage additional collaboration on novel skill proxies to make surgical training more efficient.
[ { "version": "v1", "created": "Sun, 13 Nov 2022 21:47:31 GMT" }, { "version": "v2", "created": "Thu, 30 Mar 2023 19:14:54 GMT" } ]
2023-04-03T00:00:00
[ [ "Bkheet", "Eddie", "" ], [ "D'Angelo", "Anne-Lise", "" ], [ "Goldbraikh", "Adam", "" ], [ "Laufer", "Shlomi", "" ] ]
TITLE: Using Hand Pose Estimation To Automate Open Surgery Training Feedback ABSTRACT: Purpose: This research aims to facilitate the use of state-of-the-art computer vision algorithms for the automated training of surgeons and the analysis of surgical footage. By estimating 2D hand poses, we model the movement of the practitioner's hands, and their interaction with surgical instruments, to study their potential benefit for surgical training. Methods: We leverage pre-trained models on a publicly-available hands dataset to create our own in-house dataset of 100 open surgery simulation videos with 2D hand poses. We also assess the ability of pose estimations to segment surgical videos into gestures and tool-usage segments and compare them to kinematic sensors and I3D features. Furthermore, we introduce 6 novel surgical dexterity proxies stemming from domain experts' training advice, all of which our framework can automatically detect given raw video footage. Results: State-of-the-art gesture segmentation accuracy of 88.35\% on the Open Surgery Simulation dataset is achieved with the fusion of 2D poses and I3D features from multiple angles. The introduced surgical skill proxies presented significant differences for novices compared to experts and produced actionable feedback for improvement. Conclusion: This research demonstrates the benefit of pose estimations for open surgery by analyzing their effectiveness in gesture segmentation and skill assessment. Gesture segmentation using pose estimations achieved comparable results to physical sensors while being remote and markerless. Surgical dexterity proxies that rely on pose estimation proved they can be used to work towards automated training feedback. We hope our findings encourage additional collaboration on novel skill proxies to make surgical training more efficient.
new_dataset
0.717458
2211.08447
Aiqi Jiang
Aiqi Jiang, Arkaitz Zubiaga
SexWEs: Domain-Aware Word Embeddings via Cross-lingual Semantic Specialisation for Chinese Sexism Detection in Social Media
accepted at ICWSM 2023
null
null
null
cs.CL cs.SI
http://creativecommons.org/licenses/by/4.0/
The goal of sexism detection is to mitigate negative online content targeting certain gender groups of people. However, the limited availability of labeled sexism-related datasets makes it problematic to identify online sexism for low-resource languages. In this paper, we address the task of automatic sexism detection in social media for one low-resource language -- Chinese. Rather than collecting new sexism data or building cross-lingual transfer learning models, we develop a cross-lingual domain-aware semantic specialisation system in order to make the most of existing data. Semantic specialisation is a technique for retrofitting pre-trained distributional word vectors by integrating external linguistic knowledge (such as lexico-semantic relations) into the specialised feature space. To do this, we leverage semantic resources for sexism from a high-resource language (English) to specialise pre-trained word vectors in the target language (Chinese) to inject domain knowledge. We demonstrate the benefit of our sexist word embeddings (SexWEs) specialised by our framework via intrinsic evaluation of word similarity and extrinsic evaluation of sexism detection. Compared with other specialisation approaches and Chinese baseline word vectors, our SexWEs shows an average score improvement of 0.033 and 0.064 in both intrinsic and extrinsic evaluations, respectively. The ablative results and visualisation of SexWEs also prove the effectiveness of our framework on retrofitting word vectors in low-resource languages.
[ { "version": "v1", "created": "Tue, 15 Nov 2022 19:00:20 GMT" }, { "version": "v2", "created": "Thu, 17 Nov 2022 13:05:06 GMT" }, { "version": "v3", "created": "Thu, 30 Mar 2023 21:59:01 GMT" } ]
2023-04-03T00:00:00
[ [ "Jiang", "Aiqi", "" ], [ "Zubiaga", "Arkaitz", "" ] ]
TITLE: SexWEs: Domain-Aware Word Embeddings via Cross-lingual Semantic Specialisation for Chinese Sexism Detection in Social Media ABSTRACT: The goal of sexism detection is to mitigate negative online content targeting certain gender groups of people. However, the limited availability of labeled sexism-related datasets makes it problematic to identify online sexism for low-resource languages. In this paper, we address the task of automatic sexism detection in social media for one low-resource language -- Chinese. Rather than collecting new sexism data or building cross-lingual transfer learning models, we develop a cross-lingual domain-aware semantic specialisation system in order to make the most of existing data. Semantic specialisation is a technique for retrofitting pre-trained distributional word vectors by integrating external linguistic knowledge (such as lexico-semantic relations) into the specialised feature space. To do this, we leverage semantic resources for sexism from a high-resource language (English) to specialise pre-trained word vectors in the target language (Chinese) to inject domain knowledge. We demonstrate the benefit of our sexist word embeddings (SexWEs) specialised by our framework via intrinsic evaluation of word similarity and extrinsic evaluation of sexism detection. Compared with other specialisation approaches and Chinese baseline word vectors, our SexWEs shows an average score improvement of 0.033 and 0.064 in both intrinsic and extrinsic evaluations, respectively. The ablative results and visualisation of SexWEs also prove the effectiveness of our framework on retrofitting word vectors in low-resource languages.
no_new_dataset
0.712589
2211.12277
Peng Wang
Peng Wang, Jingzhou Chen, Yuntao Qian
Semantic Guided Level-Category Hybrid Prediction Network for Hierarchical Image Classification
null
null
null
null
cs.CV
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
Hierarchical classification (HC) assigns each object with multiple labels organized into a hierarchical structure. The existing deep learning based HC methods usually predict an instance starting from the root node until a leaf node is reached. However, in the real world, images interfered by noise, occlusion, blur, or low resolution may not provide sufficient information for the classification at subordinate levels. To address this issue, we propose a novel semantic guided level-category hybrid prediction network (SGLCHPN) that can jointly perform the level and category prediction in an end-to-end manner. SGLCHPN comprises two modules: a visual transformer that extracts feature vectors from the input images, and a semantic guided cross-attention module that uses categories word embeddings as queries to guide learning category-specific representations. In order to evaluate the proposed method, we construct two new datasets in which images are at a broad range of quality and thus are labeled to different levels (depths) in the hierarchy according to their individual quality. Experimental results demonstrate the effectiveness of our proposed HC method.
[ { "version": "v1", "created": "Tue, 22 Nov 2022 13:49:10 GMT" }, { "version": "v2", "created": "Fri, 3 Mar 2023 07:18:47 GMT" }, { "version": "v3", "created": "Fri, 31 Mar 2023 08:52:12 GMT" } ]
2023-04-03T00:00:00
[ [ "Wang", "Peng", "" ], [ "Chen", "Jingzhou", "" ], [ "Qian", "Yuntao", "" ] ]
TITLE: Semantic Guided Level-Category Hybrid Prediction Network for Hierarchical Image Classification ABSTRACT: Hierarchical classification (HC) assigns each object with multiple labels organized into a hierarchical structure. The existing deep learning based HC methods usually predict an instance starting from the root node until a leaf node is reached. However, in the real world, images interfered by noise, occlusion, blur, or low resolution may not provide sufficient information for the classification at subordinate levels. To address this issue, we propose a novel semantic guided level-category hybrid prediction network (SGLCHPN) that can jointly perform the level and category prediction in an end-to-end manner. SGLCHPN comprises two modules: a visual transformer that extracts feature vectors from the input images, and a semantic guided cross-attention module that uses categories word embeddings as queries to guide learning category-specific representations. In order to evaluate the proposed method, we construct two new datasets in which images are at a broad range of quality and thus are labeled to different levels (depths) in the hierarchy according to their individual quality. Experimental results demonstrate the effectiveness of our proposed HC method.
new_dataset
0.711049
2211.16374
Gwanghyun Kim
Gwanghyun Kim and Se Young Chun
DATID-3D: Diversity-Preserved Domain Adaptation Using Text-to-Image Diffusion for 3D Generative Model
Accepted to CVPR 2023, Project page: https://gwang-kim.github.io/datid_3d/
null
null
null
cs.CV cs.AI
http://creativecommons.org/licenses/by-nc-nd/4.0/
Recent 3D generative models have achieved remarkable performance in synthesizing high resolution photorealistic images with view consistency and detailed 3D shapes, but training them for diverse domains is challenging since it requires massive training images and their camera distribution information. Text-guided domain adaptation methods have shown impressive performance on converting the 2D generative model on one domain into the models on other domains with different styles by leveraging the CLIP (Contrastive Language-Image Pre-training), rather than collecting massive datasets for those domains. However, one drawback of them is that the sample diversity in the original generative model is not well-preserved in the domain-adapted generative models due to the deterministic nature of the CLIP text encoder. Text-guided domain adaptation will be even more challenging for 3D generative models not only because of catastrophic diversity loss, but also because of inferior text-image correspondence and poor image quality. Here we propose DATID-3D, a domain adaptation method tailored for 3D generative models using text-to-image diffusion models that can synthesize diverse images per text prompt without collecting additional images and camera information for the target domain. Unlike 3D extensions of prior text-guided domain adaptation methods, our novel pipeline was able to fine-tune the state-of-the-art 3D generator of the source domain to synthesize high resolution, multi-view consistent images in text-guided targeted domains without additional data, outperforming the existing text-guided domain adaptation methods in diversity and text-image correspondence. Furthermore, we propose and demonstrate diverse 3D image manipulations such as one-shot instance-selected adaptation and single-view manipulated 3D reconstruction to fully enjoy diversity in text.
[ { "version": "v1", "created": "Tue, 29 Nov 2022 16:54:34 GMT" }, { "version": "v2", "created": "Fri, 31 Mar 2023 02:15:49 GMT" } ]
2023-04-03T00:00:00
[ [ "Kim", "Gwanghyun", "" ], [ "Chun", "Se Young", "" ] ]
TITLE: DATID-3D: Diversity-Preserved Domain Adaptation Using Text-to-Image Diffusion for 3D Generative Model ABSTRACT: Recent 3D generative models have achieved remarkable performance in synthesizing high resolution photorealistic images with view consistency and detailed 3D shapes, but training them for diverse domains is challenging since it requires massive training images and their camera distribution information. Text-guided domain adaptation methods have shown impressive performance on converting the 2D generative model on one domain into the models on other domains with different styles by leveraging the CLIP (Contrastive Language-Image Pre-training), rather than collecting massive datasets for those domains. However, one drawback of them is that the sample diversity in the original generative model is not well-preserved in the domain-adapted generative models due to the deterministic nature of the CLIP text encoder. Text-guided domain adaptation will be even more challenging for 3D generative models not only because of catastrophic diversity loss, but also because of inferior text-image correspondence and poor image quality. Here we propose DATID-3D, a domain adaptation method tailored for 3D generative models using text-to-image diffusion models that can synthesize diverse images per text prompt without collecting additional images and camera information for the target domain. Unlike 3D extensions of prior text-guided domain adaptation methods, our novel pipeline was able to fine-tune the state-of-the-art 3D generator of the source domain to synthesize high resolution, multi-view consistent images in text-guided targeted domains without additional data, outperforming the existing text-guided domain adaptation methods in diversity and text-image correspondence. Furthermore, we propose and demonstrate diverse 3D image manipulations such as one-shot instance-selected adaptation and single-view manipulated 3D reconstruction to fully enjoy diversity in text.
no_new_dataset
0.710823
2212.00338
Jiazhao Zhang
Jiazhao Zhang, Liu Dai, Fanpeng Meng, Qingnan Fan, Xuelin Chen, Kai Xu, He Wang
3D-Aware Object Goal Navigation via Simultaneous Exploration and Identification
To appear in CVPR 2023
null
null
null
cs.RO cs.CV
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
Object goal navigation (ObjectNav) in unseen environments is a fundamental task for Embodied AI. Agents in existing works learn ObjectNav policies based on 2D maps, scene graphs, or image sequences. Considering this task happens in 3D space, a 3D-aware agent can advance its ObjectNav capability via learning from fine-grained spatial information. However, leveraging 3D scene representation can be prohibitively unpractical for policy learning in this floor-level task, due to low sample efficiency and expensive computational cost. In this work, we propose a framework for the challenging 3D-aware ObjectNav based on two straightforward sub-policies. The two sub-polices, namely corner-guided exploration policy and category-aware identification policy, simultaneously perform by utilizing online fused 3D points as observation. Through extensive experiments, we show that this framework can dramatically improve the performance in ObjectNav through learning from 3D scene representation. Our framework achieves the best performance among all modular-based methods on the Matterport3D and Gibson datasets, while requiring (up to 30x) less computational cost for training.
[ { "version": "v1", "created": "Thu, 1 Dec 2022 07:55:56 GMT" }, { "version": "v2", "created": "Wed, 8 Mar 2023 16:09:04 GMT" }, { "version": "v3", "created": "Fri, 31 Mar 2023 03:49:07 GMT" } ]
2023-04-03T00:00:00
[ [ "Zhang", "Jiazhao", "" ], [ "Dai", "Liu", "" ], [ "Meng", "Fanpeng", "" ], [ "Fan", "Qingnan", "" ], [ "Chen", "Xuelin", "" ], [ "Xu", "Kai", "" ], [ "Wang", "He", "" ] ]
TITLE: 3D-Aware Object Goal Navigation via Simultaneous Exploration and Identification ABSTRACT: Object goal navigation (ObjectNav) in unseen environments is a fundamental task for Embodied AI. Agents in existing works learn ObjectNav policies based on 2D maps, scene graphs, or image sequences. Considering this task happens in 3D space, a 3D-aware agent can advance its ObjectNav capability via learning from fine-grained spatial information. However, leveraging 3D scene representation can be prohibitively unpractical for policy learning in this floor-level task, due to low sample efficiency and expensive computational cost. In this work, we propose a framework for the challenging 3D-aware ObjectNav based on two straightforward sub-policies. The two sub-polices, namely corner-guided exploration policy and category-aware identification policy, simultaneously perform by utilizing online fused 3D points as observation. Through extensive experiments, we show that this framework can dramatically improve the performance in ObjectNav through learning from 3D scene representation. Our framework achieves the best performance among all modular-based methods on the Matterport3D and Gibson datasets, while requiring (up to 30x) less computational cost for training.
no_new_dataset
0.710226
2212.04196
Yubin Wang
Cairong Zhao, Yubin Wang, Xinyang Jiang, Yifei Shen, Kaitao Song, Dongsheng Li, and Duoqian Miao
Learning Domain Invariant Prompt for Vision-Language Models
12 pages, 6 figures, 5 tables
null
null
null
cs.CV
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
Prompt learning is one of the most effective and trending ways to adapt powerful vision-language foundation models like CLIP to downstream datasets by tuning learnable prompt vectors with very few samples. However, although prompt learning achieves excellent performance over in-domain data, it still faces the major challenge of generalizing to unseen classes and domains. Some existing prompt learning methods tackle this issue by adaptively generating different prompts for different tokens or domains but neglecting the ability of learned prompts to generalize to unseen domains. In this paper, we propose a novel prompt learning paradigm that directly generates \emph{domain invariant} prompt that can be generalized to unseen domains, called MetaPrompt. Specifically, a dual-modality prompt tuning network is proposed to generate prompts for input from both image and text modalities. With a novel asymmetric contrastive loss, the representation from the original pre-trained vision-language model acts as supervision to enhance the generalization ability of the learned prompt. More importantly, we propose a meta-learning-based prompt tuning algorithm that explicitly constrains the task-specific prompt tuned for one domain or class to also achieve good performance in another domain or class. Extensive experiments on 11 datasets for base-to-new generalization and 4 datasets for domain generalization demonstrate that our method consistently and significantly outperforms existing methods.
[ { "version": "v1", "created": "Thu, 8 Dec 2022 11:23:24 GMT" }, { "version": "v2", "created": "Fri, 31 Mar 2023 04:12:25 GMT" } ]
2023-04-03T00:00:00
[ [ "Zhao", "Cairong", "" ], [ "Wang", "Yubin", "" ], [ "Jiang", "Xinyang", "" ], [ "Shen", "Yifei", "" ], [ "Song", "Kaitao", "" ], [ "Li", "Dongsheng", "" ], [ "Miao", "Duoqian", "" ] ]
TITLE: Learning Domain Invariant Prompt for Vision-Language Models ABSTRACT: Prompt learning is one of the most effective and trending ways to adapt powerful vision-language foundation models like CLIP to downstream datasets by tuning learnable prompt vectors with very few samples. However, although prompt learning achieves excellent performance over in-domain data, it still faces the major challenge of generalizing to unseen classes and domains. Some existing prompt learning methods tackle this issue by adaptively generating different prompts for different tokens or domains but neglecting the ability of learned prompts to generalize to unseen domains. In this paper, we propose a novel prompt learning paradigm that directly generates \emph{domain invariant} prompt that can be generalized to unseen domains, called MetaPrompt. Specifically, a dual-modality prompt tuning network is proposed to generate prompts for input from both image and text modalities. With a novel asymmetric contrastive loss, the representation from the original pre-trained vision-language model acts as supervision to enhance the generalization ability of the learned prompt. More importantly, we propose a meta-learning-based prompt tuning algorithm that explicitly constrains the task-specific prompt tuned for one domain or class to also achieve good performance in another domain or class. Extensive experiments on 11 datasets for base-to-new generalization and 4 datasets for domain generalization demonstrate that our method consistently and significantly outperforms existing methods.
no_new_dataset
0.709435
2212.05936
Le-Anh Tran
Le-Anh Tran, Dong-Chul Park
Encoder-Decoder Network with Guided Transmission Map: Architecture
3 pages, 2 figures, ASPAI 2022
null
null
null
cs.CV eess.IV
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
An insight into the architecture of the Encoder-Decoder Network with Guided Transmission Map (EDN-GTM), a novel and effective single image dehazing scheme, is presented in this paper. The EDN-GTM takes a conventional RGB hazy image in conjunction with the corresponding transmission map estimated by the dark channel prior (DCP) approach as inputs of the network. The EDN-GTM adopts an enhanced structure of U-Net developed for dehazing tasks and the resulting EDN-GDM has shown state-of-the-art performances on benchmark dehazing datasets in terms of PSNR and SSIM metrics. In order to give an in-depth understanding of the well-designed architecture which largely contributes to the success of the EDN-GTM, extensive experiments and analysis from selecting the core structure of the scheme to investigating advanced network designs are presented in this paper.
[ { "version": "v1", "created": "Wed, 7 Dec 2022 08:16:26 GMT" }, { "version": "v2", "created": "Fri, 31 Mar 2023 09:16:11 GMT" } ]
2023-04-03T00:00:00
[ [ "Tran", "Le-Anh", "" ], [ "Park", "Dong-Chul", "" ] ]
TITLE: Encoder-Decoder Network with Guided Transmission Map: Architecture ABSTRACT: An insight into the architecture of the Encoder-Decoder Network with Guided Transmission Map (EDN-GTM), a novel and effective single image dehazing scheme, is presented in this paper. The EDN-GTM takes a conventional RGB hazy image in conjunction with the corresponding transmission map estimated by the dark channel prior (DCP) approach as inputs of the network. The EDN-GTM adopts an enhanced structure of U-Net developed for dehazing tasks and the resulting EDN-GDM has shown state-of-the-art performances on benchmark dehazing datasets in terms of PSNR and SSIM metrics. In order to give an in-depth understanding of the well-designed architecture which largely contributes to the success of the EDN-GTM, extensive experiments and analysis from selecting the core structure of the scheme to investigating advanced network designs are presented in this paper.
no_new_dataset
0.709221
2212.12645
Austin Xu
Austin Xu, Mariya I. Vasileva, Achal Dave, Arjun Seshadri
HandsOff: Labeled Dataset Generation With No Additional Human Annotations
22 pages, 20 figures. CVPR 2023
null
null
null
cs.CV cs.LG
http://creativecommons.org/licenses/by/4.0/
Recent work leverages the expressive power of generative adversarial networks (GANs) to generate labeled synthetic datasets. These dataset generation methods often require new annotations of synthetic images, which forces practitioners to seek out annotators, curate a set of synthetic images, and ensure the quality of generated labels. We introduce the HandsOff framework, a technique capable of producing an unlimited number of synthetic images and corresponding labels after being trained on less than 50 pre-existing labeled images. Our framework avoids the practical drawbacks of prior work by unifying the field of GAN inversion with dataset generation. We generate datasets with rich pixel-wise labels in multiple challenging domains such as faces, cars, full-body human poses, and urban driving scenes. Our method achieves state-of-the-art performance in semantic segmentation, keypoint detection, and depth estimation compared to prior dataset generation approaches and transfer learning baselines. We additionally showcase its ability to address broad challenges in model development which stem from fixed, hand-annotated datasets, such as the long-tail problem in semantic segmentation. Project page: austinxu87.github.io/handsoff.
[ { "version": "v1", "created": "Sat, 24 Dec 2022 03:37:02 GMT" }, { "version": "v2", "created": "Thu, 30 Mar 2023 19:38:05 GMT" } ]
2023-04-03T00:00:00
[ [ "Xu", "Austin", "" ], [ "Vasileva", "Mariya I.", "" ], [ "Dave", "Achal", "" ], [ "Seshadri", "Arjun", "" ] ]
TITLE: HandsOff: Labeled Dataset Generation With No Additional Human Annotations ABSTRACT: Recent work leverages the expressive power of generative adversarial networks (GANs) to generate labeled synthetic datasets. These dataset generation methods often require new annotations of synthetic images, which forces practitioners to seek out annotators, curate a set of synthetic images, and ensure the quality of generated labels. We introduce the HandsOff framework, a technique capable of producing an unlimited number of synthetic images and corresponding labels after being trained on less than 50 pre-existing labeled images. Our framework avoids the practical drawbacks of prior work by unifying the field of GAN inversion with dataset generation. We generate datasets with rich pixel-wise labels in multiple challenging domains such as faces, cars, full-body human poses, and urban driving scenes. Our method achieves state-of-the-art performance in semantic segmentation, keypoint detection, and depth estimation compared to prior dataset generation approaches and transfer learning baselines. We additionally showcase its ability to address broad challenges in model development which stem from fixed, hand-annotated datasets, such as the long-tail problem in semantic segmentation. Project page: austinxu87.github.io/handsoff.
no_new_dataset
0.710245
2301.12429
Beier Zhu
Beier Zhu and Yulei Niu and Saeil Lee and Minhoe Hur and Hanwang Zhang
Debiased Fine-Tuning for Vision-language Models by Prompt Regularization
AAAI2023 accepted
null
null
null
cs.CV
http://creativecommons.org/licenses/by/4.0/
We present a new paradigm for fine-tuning large-scale visionlanguage pre-trained models on downstream task, dubbed Prompt Regularization (ProReg). Different from traditional fine-tuning which easily overfits to the downstream task data, ProReg uses the prediction by prompting the pretrained model to regularize the fine-tuning. The motivation is: by prompting the large model "a photo of a [CLASS]", the fil-lin answer is only dependent on the pretraining encyclopedic knowledge while independent of the task data distribution, which is usually biased. Specifically, given a training sample prediction during fine-tuning, we first calculate its KullbackLeibler loss of the prompt prediction and Cross-Entropy loss of the ground-truth label, and then combine them with a proposed sample-wise adaptive trade-off weight, which automatically adjusts the transfer between the pretrained and downstream domains. On various out-of-distribution benchmarks, we show the consistently strong performance of ProReg compared with conventional fine-tuning, zero-shot prompt, prompt tuning, and other state-of-the-art methods.
[ { "version": "v1", "created": "Sun, 29 Jan 2023 11:53:55 GMT" }, { "version": "v2", "created": "Fri, 31 Mar 2023 07:05:35 GMT" } ]
2023-04-03T00:00:00
[ [ "Zhu", "Beier", "" ], [ "Niu", "Yulei", "" ], [ "Lee", "Saeil", "" ], [ "Hur", "Minhoe", "" ], [ "Zhang", "Hanwang", "" ] ]
TITLE: Debiased Fine-Tuning for Vision-language Models by Prompt Regularization ABSTRACT: We present a new paradigm for fine-tuning large-scale visionlanguage pre-trained models on downstream task, dubbed Prompt Regularization (ProReg). Different from traditional fine-tuning which easily overfits to the downstream task data, ProReg uses the prediction by prompting the pretrained model to regularize the fine-tuning. The motivation is: by prompting the large model "a photo of a [CLASS]", the fil-lin answer is only dependent on the pretraining encyclopedic knowledge while independent of the task data distribution, which is usually biased. Specifically, given a training sample prediction during fine-tuning, we first calculate its KullbackLeibler loss of the prompt prediction and Cross-Entropy loss of the ground-truth label, and then combine them with a proposed sample-wise adaptive trade-off weight, which automatically adjusts the transfer between the pretrained and downstream domains. On various out-of-distribution benchmarks, we show the consistently strong performance of ProReg compared with conventional fine-tuning, zero-shot prompt, prompt tuning, and other state-of-the-art methods.
no_new_dataset
0.708036
2302.04570
Taehyung Kwon
Taehyung Kwon, Jihoon Ko, Jinhong Jung, Kijung Shin
NeuKron: Constant-Size Lossy Compression of Sparse Reorderable Matrices and Tensors
Accepted to WWW 2023 - The Web Conference 2023
null
null
null
cs.LG cs.SI
http://creativecommons.org/licenses/by/4.0/
Many real-world data are naturally represented as a sparse reorderable matrix, whose rows and columns can be arbitrarily ordered (e.g., the adjacency matrix of a bipartite graph). Storing a sparse matrix in conventional ways requires an amount of space linear in the number of non-zeros, and lossy compression of sparse matrices (e.g., Truncated SVD) typically requires an amount of space linear in the number of rows and columns. In this work, we propose NeuKron for compressing a sparse reorderable matrix into a constant-size space. NeuKron generalizes Kronecker products using a recurrent neural network with a constant number of parameters. NeuKron updates the parameters so that a given matrix is approximated by the product and reorders the rows and columns of the matrix to facilitate the approximation. The updates take time linear in the number of non-zeros in the input matrix, and the approximation of each entry can be retrieved in logarithmic time. We also extend NeuKron to compress sparse reorderable tensors (e.g. multi-layer graphs), which generalize matrices. Through experiments on ten real-world datasets, we show that NeuKron is (a) Compact: requiring up to five orders of magnitude less space than its best competitor with similar approximation errors, (b) Accurate: giving up to 10x smaller approximation error than its best competitors with similar size outputs, and (c) Scalable: successfully compressing a matrix with over 230 million non-zero entries.
[ { "version": "v1", "created": "Thu, 9 Feb 2023 11:17:34 GMT" }, { "version": "v2", "created": "Fri, 31 Mar 2023 02:43:39 GMT" } ]
2023-04-03T00:00:00
[ [ "Kwon", "Taehyung", "" ], [ "Ko", "Jihoon", "" ], [ "Jung", "Jinhong", "" ], [ "Shin", "Kijung", "" ] ]
TITLE: NeuKron: Constant-Size Lossy Compression of Sparse Reorderable Matrices and Tensors ABSTRACT: Many real-world data are naturally represented as a sparse reorderable matrix, whose rows and columns can be arbitrarily ordered (e.g., the adjacency matrix of a bipartite graph). Storing a sparse matrix in conventional ways requires an amount of space linear in the number of non-zeros, and lossy compression of sparse matrices (e.g., Truncated SVD) typically requires an amount of space linear in the number of rows and columns. In this work, we propose NeuKron for compressing a sparse reorderable matrix into a constant-size space. NeuKron generalizes Kronecker products using a recurrent neural network with a constant number of parameters. NeuKron updates the parameters so that a given matrix is approximated by the product and reorders the rows and columns of the matrix to facilitate the approximation. The updates take time linear in the number of non-zeros in the input matrix, and the approximation of each entry can be retrieved in logarithmic time. We also extend NeuKron to compress sparse reorderable tensors (e.g. multi-layer graphs), which generalize matrices. Through experiments on ten real-world datasets, we show that NeuKron is (a) Compact: requiring up to five orders of magnitude less space than its best competitor with similar approximation errors, (b) Accurate: giving up to 10x smaller approximation error than its best competitors with similar size outputs, and (c) Scalable: successfully compressing a matrix with over 230 million non-zero entries.
no_new_dataset
0.710446
2302.04694
Bjoern Andres
David Stein, Silvia Di Gregorio, Bjoern Andres
Partial Optimality in Cubic Correlation Clustering
28 pages
null
null
null
cs.DM cs.CV cs.LG
http://creativecommons.org/licenses/by/4.0/
The higher-order correlation clustering problem is an expressive model, and recently, local search heuristics have been proposed for several applications. Certifying optimality, however, is NP-hard and practically hampered already by the complexity of the problem statement. Here, we focus on establishing partial optimality conditions for the special case of complete graphs and cubic objective functions. In addition, we define and implement algorithms for testing these conditions and examine their effect numerically, on two datasets.
[ { "version": "v1", "created": "Thu, 9 Feb 2023 15:25:52 GMT" }, { "version": "v2", "created": "Fri, 31 Mar 2023 14:50:55 GMT" } ]
2023-04-03T00:00:00
[ [ "Stein", "David", "" ], [ "Di Gregorio", "Silvia", "" ], [ "Andres", "Bjoern", "" ] ]
TITLE: Partial Optimality in Cubic Correlation Clustering ABSTRACT: The higher-order correlation clustering problem is an expressive model, and recently, local search heuristics have been proposed for several applications. Certifying optimality, however, is NP-hard and practically hampered already by the complexity of the problem statement. Here, we focus on establishing partial optimality conditions for the special case of complete graphs and cubic objective functions. In addition, we define and implement algorithms for testing these conditions and examine their effect numerically, on two datasets.
no_new_dataset
0.710867
2302.09665
Zirong Chen
Zirong Chen, Issa Li, Haoxiang Zhang, Sarah Preum, John A. Stankovic, Meiyi Ma
CitySpec with Shield: A Secure Intelligent Assistant for Requirement Formalization
arXiv admin note: substantial text overlap with arXiv:2206.03132
null
null
null
cs.AI
http://creativecommons.org/licenses/by/4.0/
An increasing number of monitoring systems have been developed in smart cities to ensure that the real-time operations of a city satisfy safety and performance requirements. However, many existing city requirements are written in English with missing, inaccurate, or ambiguous information. There is a high demand for assisting city policymakers in converting human-specified requirements to machine-understandable formal specifications for monitoring systems. To tackle this limitation, we build CitySpec, the first intelligent assistant system for requirement specification in smart cities. To create CitySpec, we first collect over 1,500 real-world city requirements across different domains (e.g., transportation and energy) from over 100 cities and extract city-specific knowledge to generate a dataset of city vocabulary with 3,061 words. We also build a translation model and enhance it through requirement synthesis and develop a novel online learning framework with shielded validation. The evaluation results on real-world city requirements show that CitySpec increases the sentence-level accuracy of requirement specification from 59.02% to 86.64%, and has strong adaptability to a new city and a new domain (e.g., the F1 score for requirements in Seattle increases from 77.6% to 93.75% with online learning). After the enhancement from the shield function, CitySpec is now immune to most known textual adversarial inputs (e.g., the attack success rate of DeepWordBug after the shield function is reduced to 0% from 82.73%). We test the CitySpec with 18 participants from different domains. CitySpec shows its strong usability and adaptability to different domains, and also its robustness to malicious inputs.
[ { "version": "v1", "created": "Sun, 19 Feb 2023 20:11:06 GMT" }, { "version": "v2", "created": "Thu, 30 Mar 2023 23:25:57 GMT" } ]
2023-04-03T00:00:00
[ [ "Chen", "Zirong", "" ], [ "Li", "Issa", "" ], [ "Zhang", "Haoxiang", "" ], [ "Preum", "Sarah", "" ], [ "Stankovic", "John A.", "" ], [ "Ma", "Meiyi", "" ] ]
TITLE: CitySpec with Shield: A Secure Intelligent Assistant for Requirement Formalization ABSTRACT: An increasing number of monitoring systems have been developed in smart cities to ensure that the real-time operations of a city satisfy safety and performance requirements. However, many existing city requirements are written in English with missing, inaccurate, or ambiguous information. There is a high demand for assisting city policymakers in converting human-specified requirements to machine-understandable formal specifications for monitoring systems. To tackle this limitation, we build CitySpec, the first intelligent assistant system for requirement specification in smart cities. To create CitySpec, we first collect over 1,500 real-world city requirements across different domains (e.g., transportation and energy) from over 100 cities and extract city-specific knowledge to generate a dataset of city vocabulary with 3,061 words. We also build a translation model and enhance it through requirement synthesis and develop a novel online learning framework with shielded validation. The evaluation results on real-world city requirements show that CitySpec increases the sentence-level accuracy of requirement specification from 59.02% to 86.64%, and has strong adaptability to a new city and a new domain (e.g., the F1 score for requirements in Seattle increases from 77.6% to 93.75% with online learning). After the enhancement from the shield function, CitySpec is now immune to most known textual adversarial inputs (e.g., the attack success rate of DeepWordBug after the shield function is reduced to 0% from 82.73%). We test the CitySpec with 18 participants from different domains. CitySpec shows its strong usability and adaptability to different domains, and also its robustness to malicious inputs.
new_dataset
0.684659
2302.10272
Guang Yang
Weixun Luo, Xiaodan Xing, Guang Yang
Is Autoencoder Truly Applicable for 3D CT Super-Resolution?
ISBI 2023
null
null
null
eess.IV cs.CV
http://creativecommons.org/licenses/by/4.0/
Featured by a bottleneck structure, autoencoder (AE) and its variants have been largely applied in various medical image analysis tasks, such as segmentation, reconstruction and de-noising. Despite of their promising performances in aforementioned tasks, in this paper, we claim that AE models are not applicable to single image super-resolution (SISR) for 3D CT data. Our hypothesis is that the bottleneck architecture that resizes feature maps in AE models degrades the details of input images, thus can sabotage the performance of super-resolution. Although U-Net proposed skip connections that merge information from different levels, we claim that the degrading impact of feature resizing operations could hardly be removed by skip connections. By conducting large-scale ablation experiments and comparing the performance between models with and without the bottleneck design on a public CT lung dataset , we have discovered that AE models, including U-Net, have failed to achieve a compatible SISR result ($p<0.05$ by Student's t-test) compared to the baseline model. Our work is the first comparative study investigating the suitability of AE architecture for 3D CT SISR tasks and brings a rationale for researchers to re-think the choice of model architectures especially for 3D CT SISR tasks. The full implementation and trained models can be found at: https://github.com/Roldbach/Autoencoder-3D-CT-SISR
[ { "version": "v1", "created": "Mon, 23 Jan 2023 12:48:08 GMT" }, { "version": "v2", "created": "Fri, 31 Mar 2023 16:33:22 GMT" } ]
2023-04-03T00:00:00
[ [ "Luo", "Weixun", "" ], [ "Xing", "Xiaodan", "" ], [ "Yang", "Guang", "" ] ]
TITLE: Is Autoencoder Truly Applicable for 3D CT Super-Resolution? ABSTRACT: Featured by a bottleneck structure, autoencoder (AE) and its variants have been largely applied in various medical image analysis tasks, such as segmentation, reconstruction and de-noising. Despite of their promising performances in aforementioned tasks, in this paper, we claim that AE models are not applicable to single image super-resolution (SISR) for 3D CT data. Our hypothesis is that the bottleneck architecture that resizes feature maps in AE models degrades the details of input images, thus can sabotage the performance of super-resolution. Although U-Net proposed skip connections that merge information from different levels, we claim that the degrading impact of feature resizing operations could hardly be removed by skip connections. By conducting large-scale ablation experiments and comparing the performance between models with and without the bottleneck design on a public CT lung dataset , we have discovered that AE models, including U-Net, have failed to achieve a compatible SISR result ($p<0.05$ by Student's t-test) compared to the baseline model. Our work is the first comparative study investigating the suitability of AE architecture for 3D CT SISR tasks and brings a rationale for researchers to re-think the choice of model architectures especially for 3D CT SISR tasks. The full implementation and trained models can be found at: https://github.com/Roldbach/Autoencoder-3D-CT-SISR
no_new_dataset
0.708843
2303.11674
Jintao Guo
Jintao Guo, Na Wang, Lei Qi, Yinghuan Shi
ALOFT: A Lightweight MLP-like Architecture with Dynamic Low-frequency Transform for Domain Generalization
Accepted by CVPR2023. The code is available at https://github.com/lingeringlight/ALOFT/
null
null
null
cs.CV cs.LG
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
Domain generalization (DG) aims to learn a model that generalizes well to unseen target domains utilizing multiple source domains without re-training. Most existing DG works are based on convolutional neural networks (CNNs). However, the local operation of the convolution kernel makes the model focus too much on local representations (e.g., texture), which inherently causes the model more prone to overfit to the source domains and hampers its generalization ability. Recently, several MLP-based methods have achieved promising results in supervised learning tasks by learning global interactions among different patches of the image. Inspired by this, in this paper, we first analyze the difference between CNN and MLP methods in DG and find that MLP methods exhibit a better generalization ability because they can better capture the global representations (e.g., structure) than CNN methods. Then, based on a recent lightweight MLP method, we obtain a strong baseline that outperforms most state-of-the-art CNN-based methods. The baseline can learn global structure representations with a filter to suppress structure irrelevant information in the frequency space. Moreover, we propose a dynAmic LOw-Frequency spectrum Transform (ALOFT) that can perturb local texture features while preserving global structure features, thus enabling the filter to remove structure-irrelevant information sufficiently. Extensive experiments on four benchmarks have demonstrated that our method can achieve great performance improvement with a small number of parameters compared to SOTA CNN-based DG methods. Our code is available at https://github.com/lingeringlight/ALOFT/.
[ { "version": "v1", "created": "Tue, 21 Mar 2023 08:36:34 GMT" }, { "version": "v2", "created": "Fri, 31 Mar 2023 11:55:55 GMT" } ]
2023-04-03T00:00:00
[ [ "Guo", "Jintao", "" ], [ "Wang", "Na", "" ], [ "Qi", "Lei", "" ], [ "Shi", "Yinghuan", "" ] ]
TITLE: ALOFT: A Lightweight MLP-like Architecture with Dynamic Low-frequency Transform for Domain Generalization ABSTRACT: Domain generalization (DG) aims to learn a model that generalizes well to unseen target domains utilizing multiple source domains without re-training. Most existing DG works are based on convolutional neural networks (CNNs). However, the local operation of the convolution kernel makes the model focus too much on local representations (e.g., texture), which inherently causes the model more prone to overfit to the source domains and hampers its generalization ability. Recently, several MLP-based methods have achieved promising results in supervised learning tasks by learning global interactions among different patches of the image. Inspired by this, in this paper, we first analyze the difference between CNN and MLP methods in DG and find that MLP methods exhibit a better generalization ability because they can better capture the global representations (e.g., structure) than CNN methods. Then, based on a recent lightweight MLP method, we obtain a strong baseline that outperforms most state-of-the-art CNN-based methods. The baseline can learn global structure representations with a filter to suppress structure irrelevant information in the frequency space. Moreover, we propose a dynAmic LOw-Frequency spectrum Transform (ALOFT) that can perturb local texture features while preserving global structure features, thus enabling the filter to remove structure-irrelevant information sufficiently. Extensive experiments on four benchmarks have demonstrated that our method can achieve great performance improvement with a small number of parameters compared to SOTA CNN-based DG methods. Our code is available at https://github.com/lingeringlight/ALOFT/.
no_new_dataset
0.711074
2303.14753
Andreas Kirsch
Andreas Kirsch
Does "Deep Learning on a Data Diet" reproduce? Overall yes, but GraNd at Initialization does not
5 pages
null
null
null
cs.LG cs.CV
http://creativecommons.org/licenses/by/4.0/
The paper 'Deep Learning on a Data Diet' by Paul et al. (2021) introduces two innovative metrics for pruning datasets during the training of neural networks. While we are able to replicate the results for the EL2N score at epoch 20, the same cannot be said for the GraNd score at initialization. The GraNd scores later in training provide useful pruning signals, however. The GraNd score at initialization calculates the average gradient norm of an input sample across multiple randomly initialized models before any training has taken place. Our analysis reveals a strong correlation between the GraNd score at initialization and the input norm of a sample, suggesting that the latter could have been a cheap new baseline for data pruning. Unfortunately, neither the GraNd score at initialization nor the input norm surpasses random pruning in performance. This contradicts one of the findings in Paul et al. (2021). We were unable to reproduce their CIFAR-10 results using both an updated version of the original JAX repository and in a newly implemented PyTorch codebase. An investigation of the underlying JAX/FLAX code from 2021 surfaced a bug in the checkpoint restoring code that was fixed in April 2021 (https://github.com/google/flax/commit/28fbd95500f4bf2f9924d2560062fa50e919b1a5).
[ { "version": "v1", "created": "Sun, 26 Mar 2023 15:13:19 GMT" } ]
2023-04-03T00:00:00
[ [ "Kirsch", "Andreas", "" ] ]
TITLE: Does "Deep Learning on a Data Diet" reproduce? Overall yes, but GraNd at Initialization does not ABSTRACT: The paper 'Deep Learning on a Data Diet' by Paul et al. (2021) introduces two innovative metrics for pruning datasets during the training of neural networks. While we are able to replicate the results for the EL2N score at epoch 20, the same cannot be said for the GraNd score at initialization. The GraNd scores later in training provide useful pruning signals, however. The GraNd score at initialization calculates the average gradient norm of an input sample across multiple randomly initialized models before any training has taken place. Our analysis reveals a strong correlation between the GraNd score at initialization and the input norm of a sample, suggesting that the latter could have been a cheap new baseline for data pruning. Unfortunately, neither the GraNd score at initialization nor the input norm surpasses random pruning in performance. This contradicts one of the findings in Paul et al. (2021). We were unable to reproduce their CIFAR-10 results using both an updated version of the original JAX repository and in a newly implemented PyTorch codebase. An investigation of the underlying JAX/FLAX code from 2021 surfaced a bug in the checkpoint restoring code that was fixed in April 2021 (https://github.com/google/flax/commit/28fbd95500f4bf2f9924d2560062fa50e919b1a5).
no_new_dataset
0.706361
2303.15616
Xuyang Shen
Xuyang Shen and Dong Li and Jinxing Zhou and Zhen Qin and Bowen He and Xiaodong Han and Aixuan Li and Yuchao Dai and Lingpeng Kong and Meng Wang and Yu Qiao and Yiran Zhong
Fine-grained Audible Video Description
accepted to CVPR 2023, Xuyang Shen, Dong Li and Jinxing Zhou contribute equally, code link: github.com/OpenNLPLab/FAVDBench, dataset link: www.avlbench.opennlplab.cn
null
null
17
cs.CV
http://creativecommons.org/publicdomain/zero/1.0/
We explore a new task for audio-visual-language modeling called fine-grained audible video description (FAVD). It aims to provide detailed textual descriptions for the given audible videos, including the appearance and spatial locations of each object, the actions of moving objects, and the sounds in videos. Existing visual-language modeling tasks often concentrate on visual cues in videos while undervaluing the language and audio modalities. On the other hand, FAVD requires not only audio-visual-language modeling skills but also paragraph-level language generation abilities. We construct the first fine-grained audible video description benchmark (FAVDBench) to facilitate this research. For each video clip, we first provide a one-sentence summary of the video, ie, the caption, followed by 4-6 sentences describing the visual details and 1-2 audio-related descriptions at the end. The descriptions are provided in both English and Chinese. We create two new metrics for this task: an EntityScore to gauge the completeness of entities in the visual descriptions, and an AudioScore to assess the audio descriptions. As a preliminary approach to this task, we propose an audio-visual-language transformer that extends existing video captioning model with an additional audio branch. We combine the masked language modeling and auto-regressive language modeling losses to optimize our model so that it can produce paragraph-level descriptions. We illustrate the efficiency of our model in audio-visual-language modeling by evaluating it against the proposed benchmark using both conventional captioning metrics and our proposed metrics. We further put our benchmark to the test in video generation models, demonstrating that employing fine-grained video descriptions can create more intricate videos than using captions.
[ { "version": "v1", "created": "Mon, 27 Mar 2023 22:03:48 GMT" } ]
2023-04-03T00:00:00
[ [ "Shen", "Xuyang", "" ], [ "Li", "Dong", "" ], [ "Zhou", "Jinxing", "" ], [ "Qin", "Zhen", "" ], [ "He", "Bowen", "" ], [ "Han", "Xiaodong", "" ], [ "Li", "Aixuan", "" ], [ "Dai", "Yuchao", "" ], [ "Kong", "Lingpeng", "" ], [ "Wang", "Meng", "" ], [ "Qiao", "Yu", "" ], [ "Zhong", "Yiran", "" ] ]
TITLE: Fine-grained Audible Video Description ABSTRACT: We explore a new task for audio-visual-language modeling called fine-grained audible video description (FAVD). It aims to provide detailed textual descriptions for the given audible videos, including the appearance and spatial locations of each object, the actions of moving objects, and the sounds in videos. Existing visual-language modeling tasks often concentrate on visual cues in videos while undervaluing the language and audio modalities. On the other hand, FAVD requires not only audio-visual-language modeling skills but also paragraph-level language generation abilities. We construct the first fine-grained audible video description benchmark (FAVDBench) to facilitate this research. For each video clip, we first provide a one-sentence summary of the video, ie, the caption, followed by 4-6 sentences describing the visual details and 1-2 audio-related descriptions at the end. The descriptions are provided in both English and Chinese. We create two new metrics for this task: an EntityScore to gauge the completeness of entities in the visual descriptions, and an AudioScore to assess the audio descriptions. As a preliminary approach to this task, we propose an audio-visual-language transformer that extends existing video captioning model with an additional audio branch. We combine the masked language modeling and auto-regressive language modeling losses to optimize our model so that it can produce paragraph-level descriptions. We illustrate the efficiency of our model in audio-visual-language modeling by evaluating it against the proposed benchmark using both conventional captioning metrics and our proposed metrics. We further put our benchmark to the test in video generation models, demonstrating that employing fine-grained video descriptions can create more intricate videos than using captions.
no_new_dataset
0.707967
2303.16066
Chenxi Huang
Chenxi Huang and Liang Xie and Yibo Yang and Wenxiao Wang and Binbin Lin and Deng Cai
Neural Collapse Inspired Federated Learning with Non-iid Data
11 pages, 5 figures
null
null
null
cs.LG cs.CV
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
One of the challenges in federated learning is the non-independent and identically distributed (non-iid) characteristics between heterogeneous devices, which cause significant differences in local updates and affect the performance of the central server. Although many studies have been proposed to address this challenge, they only focus on local training and aggregation processes to smooth the changes and fail to achieve high performance with deep learning models. Inspired by the phenomenon of neural collapse, we force each client to be optimized toward an optimal global structure for classification. Specifically, we initialize it as a random simplex Equiangular Tight Frame (ETF) and fix it as the unit optimization target of all clients during the local updating. After guaranteeing all clients are learning to converge to the global optimum, we propose to add a global memory vector for each category to remedy the parameter fluctuation caused by the bias of the intra-class condition distribution among clients. Our experimental results show that our method can improve the performance with faster convergence speed on different-size datasets.
[ { "version": "v1", "created": "Mon, 27 Mar 2023 05:29:53 GMT" }, { "version": "v2", "created": "Fri, 31 Mar 2023 09:54:13 GMT" } ]
2023-04-03T00:00:00
[ [ "Huang", "Chenxi", "" ], [ "Xie", "Liang", "" ], [ "Yang", "Yibo", "" ], [ "Wang", "Wenxiao", "" ], [ "Lin", "Binbin", "" ], [ "Cai", "Deng", "" ] ]
TITLE: Neural Collapse Inspired Federated Learning with Non-iid Data ABSTRACT: One of the challenges in federated learning is the non-independent and identically distributed (non-iid) characteristics between heterogeneous devices, which cause significant differences in local updates and affect the performance of the central server. Although many studies have been proposed to address this challenge, they only focus on local training and aggregation processes to smooth the changes and fail to achieve high performance with deep learning models. Inspired by the phenomenon of neural collapse, we force each client to be optimized toward an optimal global structure for classification. Specifically, we initialize it as a random simplex Equiangular Tight Frame (ETF) and fix it as the unit optimization target of all clients during the local updating. After guaranteeing all clients are learning to converge to the global optimum, we propose to add a global memory vector for each category to remedy the parameter fluctuation caused by the bias of the intra-class condition distribution among clients. Our experimental results show that our method can improve the performance with faster convergence speed on different-size datasets.
no_new_dataset
0.709573
2303.17611
Mohammed Daoudi
Yujin Wu, Mohamed Daoudi, Ali Amad
Transformer-based Self-supervised Multimodal Representation Learning for Wearable Emotion Recognition
Accepted IEEE Transactions On Affective Computing
null
null
null
cs.HC cs.AI cs.SD eess.AS
http://creativecommons.org/licenses/by/4.0/
Recently, wearable emotion recognition based on peripheral physiological signals has drawn massive attention due to its less invasive nature and its applicability in real-life scenarios. However, how to effectively fuse multimodal data remains a challenging problem. Moreover, traditional fully-supervised based approaches suffer from overfitting given limited labeled data. To address the above issues, we propose a novel self-supervised learning (SSL) framework for wearable emotion recognition, where efficient multimodal fusion is realized with temporal convolution-based modality-specific encoders and a transformer-based shared encoder, capturing both intra-modal and inter-modal correlations. Extensive unlabeled data is automatically assigned labels by five signal transforms, and the proposed SSL model is pre-trained with signal transformation recognition as a pretext task, allowing the extraction of generalized multimodal representations for emotion-related downstream tasks. For evaluation, the proposed SSL model was first pre-trained on a large-scale self-collected physiological dataset and the resulting encoder was subsequently frozen or fine-tuned on three public supervised emotion recognition datasets. Ultimately, our SSL-based method achieved state-of-the-art results in various emotion classification tasks. Meanwhile, the proposed model proved to be more accurate and robust compared to fully-supervised methods on low data regimes.
[ { "version": "v1", "created": "Wed, 29 Mar 2023 19:45:55 GMT" } ]
2023-04-03T00:00:00
[ [ "Wu", "Yujin", "" ], [ "Daoudi", "Mohamed", "" ], [ "Amad", "Ali", "" ] ]
TITLE: Transformer-based Self-supervised Multimodal Representation Learning for Wearable Emotion Recognition ABSTRACT: Recently, wearable emotion recognition based on peripheral physiological signals has drawn massive attention due to its less invasive nature and its applicability in real-life scenarios. However, how to effectively fuse multimodal data remains a challenging problem. Moreover, traditional fully-supervised based approaches suffer from overfitting given limited labeled data. To address the above issues, we propose a novel self-supervised learning (SSL) framework for wearable emotion recognition, where efficient multimodal fusion is realized with temporal convolution-based modality-specific encoders and a transformer-based shared encoder, capturing both intra-modal and inter-modal correlations. Extensive unlabeled data is automatically assigned labels by five signal transforms, and the proposed SSL model is pre-trained with signal transformation recognition as a pretext task, allowing the extraction of generalized multimodal representations for emotion-related downstream tasks. For evaluation, the proposed SSL model was first pre-trained on a large-scale self-collected physiological dataset and the resulting encoder was subsequently frozen or fine-tuned on three public supervised emotion recognition datasets. Ultimately, our SSL-based method achieved state-of-the-art results in various emotion classification tasks. Meanwhile, the proposed model proved to be more accurate and robust compared to fully-supervised methods on low data regimes.
no_new_dataset
0.71039
2303.17617
Ioannis Kontopoulos Dr.
Ioannis Kontopoulos, Antonios Makris, Konstantinos Tserpes, Theodora Varvarigou
An evaluation of time series forecasting models on water consumption data: A case study of Greece
null
null
null
null
cs.LG
http://creativecommons.org/licenses/by-nc-nd/4.0/
In recent years, the increased urbanization and industrialization has led to a rising water demand and resources, thus increasing the gap between demand and supply. Proper water distribution and forecasting of water consumption are key factors in mitigating the imbalance of supply and demand by improving operations, planning and management of water resources. To this end, in this paper, several well-known forecasting algorithms are evaluated over time series, water consumption data from Greece, a country with diverse socio-economic and urbanization issues. The forecasting algorithms are evaluated on a real-world dataset provided by the Water Supply and Sewerage Company of Greece revealing key insights about each algorithm and its use.
[ { "version": "v1", "created": "Thu, 30 Mar 2023 09:09:41 GMT" } ]
2023-04-03T00:00:00
[ [ "Kontopoulos", "Ioannis", "" ], [ "Makris", "Antonios", "" ], [ "Tserpes", "Konstantinos", "" ], [ "Varvarigou", "Theodora", "" ] ]
TITLE: An evaluation of time series forecasting models on water consumption data: A case study of Greece ABSTRACT: In recent years, the increased urbanization and industrialization has led to a rising water demand and resources, thus increasing the gap between demand and supply. Proper water distribution and forecasting of water consumption are key factors in mitigating the imbalance of supply and demand by improving operations, planning and management of water resources. To this end, in this paper, several well-known forecasting algorithms are evaluated over time series, water consumption data from Greece, a country with diverse socio-economic and urbanization issues. The forecasting algorithms are evaluated on a real-world dataset provided by the Water Supply and Sewerage Company of Greece revealing key insights about each algorithm and its use.
no_new_dataset
0.711205
2303.17619
Pooja Prajod
Pooja Prajod, Matteo Lavit Nicora, Matteo Malosio, Elisabeth Andr\'e
Gaze-based Attention Recognition for Human-Robot Collaboration
Accepted to PETRA 2023
null
null
null
cs.HC cs.AI cs.CV cs.RO
http://creativecommons.org/licenses/by-sa/4.0/
Attention (and distraction) recognition is a key factor in improving human-robot collaboration. We present an assembly scenario where a human operator and a cobot collaborate equally to piece together a gearbox. The setup provides multiple opportunities for the cobot to adapt its behavior depending on the operator's attention, which can improve the collaboration experience and reduce psychological strain. As a first step, we recognize the areas in the workspace that the human operator is paying attention to, and consequently, detect when the operator is distracted. We propose a novel deep-learning approach to develop an attention recognition model. First, we train a convolutional neural network to estimate the gaze direction using a publicly available image dataset. Then, we use transfer learning with a small dataset to map the gaze direction onto pre-defined areas of interest. Models trained using this approach performed very well in leave-one-subject-out evaluation on the small dataset. We performed an additional validation of our models using the video snippets collected from participants working as an operator in the presented assembly scenario. Although the recall for the Distracted class was lower in this case, the models performed well in recognizing the areas the operator paid attention to. To the best of our knowledge, this is the first work that validated an attention recognition model using data from a setting that mimics industrial human-robot collaboration. Our findings highlight the need for validation of attention recognition solutions in such full-fledged, non-guided scenarios.
[ { "version": "v1", "created": "Thu, 30 Mar 2023 11:55:38 GMT" } ]
2023-04-03T00:00:00
[ [ "Prajod", "Pooja", "" ], [ "Nicora", "Matteo Lavit", "" ], [ "Malosio", "Matteo", "" ], [ "André", "Elisabeth", "" ] ]
TITLE: Gaze-based Attention Recognition for Human-Robot Collaboration ABSTRACT: Attention (and distraction) recognition is a key factor in improving human-robot collaboration. We present an assembly scenario where a human operator and a cobot collaborate equally to piece together a gearbox. The setup provides multiple opportunities for the cobot to adapt its behavior depending on the operator's attention, which can improve the collaboration experience and reduce psychological strain. As a first step, we recognize the areas in the workspace that the human operator is paying attention to, and consequently, detect when the operator is distracted. We propose a novel deep-learning approach to develop an attention recognition model. First, we train a convolutional neural network to estimate the gaze direction using a publicly available image dataset. Then, we use transfer learning with a small dataset to map the gaze direction onto pre-defined areas of interest. Models trained using this approach performed very well in leave-one-subject-out evaluation on the small dataset. We performed an additional validation of our models using the video snippets collected from participants working as an operator in the presented assembly scenario. Although the recall for the Distracted class was lower in this case, the models performed well in recognizing the areas the operator paid attention to. To the best of our knowledge, this is the first work that validated an attention recognition model using data from a setting that mimics industrial human-robot collaboration. Our findings highlight the need for validation of attention recognition solutions in such full-fledged, non-guided scenarios.
no_new_dataset
0.708231
2303.17636
Dominik Bati\'c
Dominik Bati\'c, Felix Holm, Ege \"Ozsoy, Tobias Czempiel, Nassir Navab
Whether and When does Endoscopy Domain Pretraining Make Sense?
null
null
null
null
cs.CV
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
Automated endoscopy video analysis is a challenging task in medical computer vision, with the primary objective of assisting surgeons during procedures. The difficulty arises from the complexity of surgical scenes and the lack of a sufficient amount of annotated data. In recent years, large-scale pretraining has shown great success in natural language processing and computer vision communities. These approaches reduce the need for annotated data, which is always a concern in the medical domain. However, most works on endoscopic video understanding use models pretrained on natural images, creating a domain gap between pretraining and finetuning. In this work, we investigate the need for endoscopy domain-specific pretraining based on downstream objectives. To this end, we first collect Endo700k, the largest publicly available corpus of endoscopic images, extracted from nine public Minimally Invasive Surgery (MIS) datasets. Endo700k comprises more than 700,000 unannotated raw images. Next, we introduce EndoViT, an endoscopy pretrained Vision Transformer (ViT). Through ablations, we demonstrate that domain-specific pretraining is particularly beneficial for more complex downstream tasks, such as Action Triplet Detection, and less effective and even unnecessary for simpler tasks, such as Surgical Phase Recognition. We will release both our code and pretrained models upon acceptance to facilitate further research in this direction.
[ { "version": "v1", "created": "Thu, 30 Mar 2023 18:01:26 GMT" } ]
2023-04-03T00:00:00
[ [ "Batić", "Dominik", "" ], [ "Holm", "Felix", "" ], [ "Özsoy", "Ege", "" ], [ "Czempiel", "Tobias", "" ], [ "Navab", "Nassir", "" ] ]
TITLE: Whether and When does Endoscopy Domain Pretraining Make Sense? ABSTRACT: Automated endoscopy video analysis is a challenging task in medical computer vision, with the primary objective of assisting surgeons during procedures. The difficulty arises from the complexity of surgical scenes and the lack of a sufficient amount of annotated data. In recent years, large-scale pretraining has shown great success in natural language processing and computer vision communities. These approaches reduce the need for annotated data, which is always a concern in the medical domain. However, most works on endoscopic video understanding use models pretrained on natural images, creating a domain gap between pretraining and finetuning. In this work, we investigate the need for endoscopy domain-specific pretraining based on downstream objectives. To this end, we first collect Endo700k, the largest publicly available corpus of endoscopic images, extracted from nine public Minimally Invasive Surgery (MIS) datasets. Endo700k comprises more than 700,000 unannotated raw images. Next, we introduce EndoViT, an endoscopy pretrained Vision Transformer (ViT). Through ablations, we demonstrate that domain-specific pretraining is particularly beneficial for more complex downstream tasks, such as Action Triplet Detection, and less effective and even unnecessary for simpler tasks, such as Surgical Phase Recognition. We will release both our code and pretrained models upon acceptance to facilitate further research in this direction.
no_new_dataset
0.641759
2303.17644
Rhydian Windsor
Rhydian Windsor, Amir Jamaludin, Timor Kadir, Andrew Zisserman
Vision-Language Modelling For Radiological Imaging and Reports In The Low Data Regime
Accepted to MIDL 2023
null
null
null
cs.CV
http://creativecommons.org/licenses/by/4.0/
This paper explores training medical vision-language models (VLMs) -- where the visual and language inputs are embedded into a common space -- with a particular focus on scenarios where training data is limited, as is often the case in clinical datasets. We explore several candidate methods to improve low-data performance, including: (i) adapting generic pre-trained models to novel image and text domains (i.e. medical imaging and reports) via unimodal self-supervision; (ii) using local (e.g. GLoRIA) & global (e.g. InfoNCE) contrastive loss functions as well as a combination of the two; (iii) extra supervision during VLM training, via: (a) image- and text-only self-supervision, and (b) creating additional positive image-text pairs for training through augmentation and nearest-neighbour search. Using text-to-image retrieval as a benchmark, we evaluate the performance of these methods with variable sized training datasets of paired chest X-rays and radiological reports. Combined, they significantly improve retrieval compared to fine-tuning CLIP, roughly equivalent to training with the data. A similar pattern is found in the downstream task classification of CXR-related conditions with our method outperforming CLIP and also BioVIL, a strong CXR VLM benchmark, in the zero-shot and linear probing settings. We conclude with a set of recommendations for researchers aiming to train vision-language models on other medical imaging modalities when training data is scarce. To facilitate further research, we will make our code and models publicly available.
[ { "version": "v1", "created": "Thu, 30 Mar 2023 18:20:00 GMT" } ]
2023-04-03T00:00:00
[ [ "Windsor", "Rhydian", "" ], [ "Jamaludin", "Amir", "" ], [ "Kadir", "Timor", "" ], [ "Zisserman", "Andrew", "" ] ]
TITLE: Vision-Language Modelling For Radiological Imaging and Reports In The Low Data Regime ABSTRACT: This paper explores training medical vision-language models (VLMs) -- where the visual and language inputs are embedded into a common space -- with a particular focus on scenarios where training data is limited, as is often the case in clinical datasets. We explore several candidate methods to improve low-data performance, including: (i) adapting generic pre-trained models to novel image and text domains (i.e. medical imaging and reports) via unimodal self-supervision; (ii) using local (e.g. GLoRIA) & global (e.g. InfoNCE) contrastive loss functions as well as a combination of the two; (iii) extra supervision during VLM training, via: (a) image- and text-only self-supervision, and (b) creating additional positive image-text pairs for training through augmentation and nearest-neighbour search. Using text-to-image retrieval as a benchmark, we evaluate the performance of these methods with variable sized training datasets of paired chest X-rays and radiological reports. Combined, they significantly improve retrieval compared to fine-tuning CLIP, roughly equivalent to training with the data. A similar pattern is found in the downstream task classification of CXR-related conditions with our method outperforming CLIP and also BioVIL, a strong CXR VLM benchmark, in the zero-shot and linear probing settings. We conclude with a set of recommendations for researchers aiming to train vision-language models on other medical imaging modalities when training data is scarce. To facilitate further research, we will make our code and models publicly available.
no_new_dataset
0.710597
2303.17647
Danyang Liu
Danyang Liu, Frank Keller
Detecting and Grounding Important Characters in Visual Stories
AAAI 2023
null
null
null
cs.CL cs.CV
http://creativecommons.org/licenses/by/4.0/
Characters are essential to the plot of any story. Establishing the characters before writing a story can improve the clarity of the plot and the overall flow of the narrative. However, previous work on visual storytelling tends to focus on detecting objects in images and discovering relationships between them. In this approach, characters are not distinguished from other objects when they are fed into the generation pipeline. The result is a coherent sequence of events rather than a character-centric story. In order to address this limitation, we introduce the VIST-Character dataset, which provides rich character-centric annotations, including visual and textual co-reference chains and importance ratings for characters. Based on this dataset, we propose two new tasks: important character detection and character grounding in visual stories. For both tasks, we develop simple, unsupervised models based on distributional similarity and pre-trained vision-and-language models. Our new dataset, together with these models, can serve as the foundation for subsequent work on analysing and generating stories from a character-centric perspective.
[ { "version": "v1", "created": "Thu, 30 Mar 2023 18:24:06 GMT" } ]
2023-04-03T00:00:00
[ [ "Liu", "Danyang", "" ], [ "Keller", "Frank", "" ] ]
TITLE: Detecting and Grounding Important Characters in Visual Stories ABSTRACT: Characters are essential to the plot of any story. Establishing the characters before writing a story can improve the clarity of the plot and the overall flow of the narrative. However, previous work on visual storytelling tends to focus on detecting objects in images and discovering relationships between them. In this approach, characters are not distinguished from other objects when they are fed into the generation pipeline. The result is a coherent sequence of events rather than a character-centric story. In order to address this limitation, we introduce the VIST-Character dataset, which provides rich character-centric annotations, including visual and textual co-reference chains and importance ratings for characters. Based on this dataset, we propose two new tasks: important character detection and character grounding in visual stories. For both tasks, we develop simple, unsupervised models based on distributional similarity and pre-trained vision-and-language models. Our new dataset, together with these models, can serve as the foundation for subsequent work on analysing and generating stories from a character-centric perspective.
new_dataset
0.716993
2303.17661
Muntabir Hasan Choudhury
Muntabir Hasan Choudhury, Lamia Salsabil, Himarsha R. Jayanetti, Jian Wu, William A. Ingram, Edward A. Fox
MetaEnhance: Metadata Quality Improvement for Electronic Theses and Dissertations of University Libraries
7 pages, 3 tables, and 1 figure. Accepted by 2023 ACM/IEEE Joint Conference on Digital Libraries (JCDL '23) as a short paper
null
null
null
cs.DL cs.AI cs.LG
http://creativecommons.org/licenses/by-nc-sa/4.0/
Metadata quality is crucial for digital objects to be discovered through digital library interfaces. However, due to various reasons, the metadata of digital objects often exhibits incomplete, inconsistent, and incorrect values. We investigate methods to automatically detect, correct, and canonicalize scholarly metadata, using seven key fields of electronic theses and dissertations (ETDs) as a case study. We propose MetaEnhance, a framework that utilizes state-of-the-art artificial intelligence methods to improve the quality of these fields. To evaluate MetaEnhance, we compiled a metadata quality evaluation benchmark containing 500 ETDs, by combining subsets sampled using multiple criteria. We tested MetaEnhance on this benchmark and found that the proposed methods achieved nearly perfect F1-scores in detecting errors and F1-scores in correcting errors ranging from 0.85 to 1.00 for five of seven fields.
[ { "version": "v1", "created": "Thu, 30 Mar 2023 18:56:42 GMT" } ]
2023-04-03T00:00:00
[ [ "Choudhury", "Muntabir Hasan", "" ], [ "Salsabil", "Lamia", "" ], [ "Jayanetti", "Himarsha R.", "" ], [ "Wu", "Jian", "" ], [ "Ingram", "William A.", "" ], [ "Fox", "Edward A.", "" ] ]
TITLE: MetaEnhance: Metadata Quality Improvement for Electronic Theses and Dissertations of University Libraries ABSTRACT: Metadata quality is crucial for digital objects to be discovered through digital library interfaces. However, due to various reasons, the metadata of digital objects often exhibits incomplete, inconsistent, and incorrect values. We investigate methods to automatically detect, correct, and canonicalize scholarly metadata, using seven key fields of electronic theses and dissertations (ETDs) as a case study. We propose MetaEnhance, a framework that utilizes state-of-the-art artificial intelligence methods to improve the quality of these fields. To evaluate MetaEnhance, we compiled a metadata quality evaluation benchmark containing 500 ETDs, by combining subsets sampled using multiple criteria. We tested MetaEnhance on this benchmark and found that the proposed methods achieved nearly perfect F1-scores in detecting errors and F1-scores in correcting errors ranging from 0.85 to 1.00 for five of seven fields.
new_dataset
0.714653
2303.17688
Aiyu Cui
Aiyu Cui, Sen He, Tao Xiang, Antoine Toisoul
Learning Garment DensePose for Robust Warping in Virtual Try-On
6 pages
null
null
null
cs.CV
http://creativecommons.org/licenses/by/4.0/
Virtual try-on, i.e making people virtually try new garments, is an active research area in computer vision with great commercial applications. Current virtual try-on methods usually work in a two-stage pipeline. First, the garment image is warped on the person's pose using a flow estimation network. Then in the second stage, the warped garment is fused with the person image to render a new try-on image. Unfortunately, such methods are heavily dependent on the quality of the garment warping which often fails when dealing with hard poses (e.g., a person lifting or crossing arms). In this work, we propose a robust warping method for virtual try-on based on a learned garment DensePose which has a direct correspondence with the person's DensePose. Due to the lack of annotated data, we show how to leverage an off-the-shelf person DensePose model and a pretrained flow model to learn the garment DensePose in a weakly supervised manner. The garment DensePose allows a robust warping to any person's pose without any additional computation. Our method achieves the state-of-the-art equivalent on virtual try-on benchmarks and shows warping robustness on in-the-wild person images with hard poses, making it more suited for real-world virtual try-on applications.
[ { "version": "v1", "created": "Thu, 30 Mar 2023 20:02:29 GMT" } ]
2023-04-03T00:00:00
[ [ "Cui", "Aiyu", "" ], [ "He", "Sen", "" ], [ "Xiang", "Tao", "" ], [ "Toisoul", "Antoine", "" ] ]
TITLE: Learning Garment DensePose for Robust Warping in Virtual Try-On ABSTRACT: Virtual try-on, i.e making people virtually try new garments, is an active research area in computer vision with great commercial applications. Current virtual try-on methods usually work in a two-stage pipeline. First, the garment image is warped on the person's pose using a flow estimation network. Then in the second stage, the warped garment is fused with the person image to render a new try-on image. Unfortunately, such methods are heavily dependent on the quality of the garment warping which often fails when dealing with hard poses (e.g., a person lifting or crossing arms). In this work, we propose a robust warping method for virtual try-on based on a learned garment DensePose which has a direct correspondence with the person's DensePose. Due to the lack of annotated data, we show how to leverage an off-the-shelf person DensePose model and a pretrained flow model to learn the garment DensePose in a weakly supervised manner. The garment DensePose allows a robust warping to any person's pose without any additional computation. Our method achieves the state-of-the-art equivalent on virtual try-on benchmarks and shows warping robustness on in-the-wild person images with hard poses, making it more suited for real-world virtual try-on applications.
no_new_dataset
0.711406
2303.17719
Matthias Eisenmann
Matthias Eisenmann, Annika Reinke, Vivienn Weru, Minu Dietlinde Tizabi, Fabian Isensee, Tim J. Adler, Sharib Ali, Vincent Andrearczyk, Marc Aubreville, Ujjwal Baid, Spyridon Bakas, Niranjan Balu, Sophia Bano, Jorge Bernal, Sebastian Bodenstedt, Alessandro Casella, Veronika Cheplygina, Marie Daum, Marleen de Bruijne, Adrien Depeursinge, Reuben Dorent, Jan Egger, David G. Ellis, Sandy Engelhardt, Melanie Ganz, Noha Ghatwary, Gabriel Girard, Patrick Godau, Anubha Gupta, Lasse Hansen, Kanako Harada, Mattias Heinrich, Nicholas Heller, Alessa Hering, Arnaud Huaulm\'e, Pierre Jannin, Ali Emre Kavur, Old\v{r}ich Kodym, Michal Kozubek, Jianning Li, Hongwei Li, Jun Ma, Carlos Mart\'in-Isla, Bjoern Menze, Alison Noble, Valentin Oreiller, Nicolas Padoy, Sarthak Pati, Kelly Payette, Tim R\"adsch, Jonathan Rafael-Pati\~no, Vivek Singh Bawa, Stefanie Speidel, Carole H. Sudre, Kimberlin van Wijnen, Martin Wagner, Donglai Wei, Amine Yamlahi, Moi Hoon Yap, Chun Yuan, Maximilian Zenk, Aneeq Zia, David Zimmerer, Dogu Baran Aydogan, Binod Bhattarai, Louise Bloch, Raphael Br\"ungel, Jihoon Cho, Chanyeol Choi, Qi Dou, Ivan Ezhov, Christoph M. Friedrich, Clifton Fuller, Rebati Raman Gaire, Adrian Galdran, \'Alvaro Garc\'ia Faura, Maria Grammatikopoulou, SeulGi Hong, Mostafa Jahanifar, Ikbeom Jang, Abdolrahim Kadkhodamohammadi, Inha Kang, Florian Kofler, Satoshi Kondo, Hugo Kuijf, Mingxing Li, Minh Huan Luu, Toma\v{z} Martin\v{c}i\v{c}, Pedro Morais, Mohamed A. Naser, Bruno Oliveira, David Owen, Subeen Pang, Jinah Park, Sung-Hong Park, Szymon P{\l}otka, Elodie Puybareau, Nasir Rajpoot, Kanghyun Ryu, Numan Saeed, Adam Shephard, Pengcheng Shi, Dejan \v{S}tepec, Ronast Subedi, Guillaume Tochon, Helena R. Torres, Helene Urien, Jo\~ao L. Vila\c{c}a, Kareem Abdul Wahid, Haojie Wang, Jiacheng Wang, Liansheng Wang, Xiyue Wang, Benedikt Wiestler, Marek Wodzinski, Fangfang Xia, Juanying Xie, Zhiwei Xiong, Sen Yang, Yanwu Yang, Zixuan Zhao, Klaus Maier-Hein, Paul F. J\"ager, Annette Kopp-Schneider, and Lena Maier-Hein
Why is the winner the best?
accepted to CVPR 2023
null
null
null
cs.CV cs.LG
http://creativecommons.org/licenses/by/4.0/
International benchmarking competitions have become fundamental for the comparative performance assessment of image analysis methods. However, little attention has been given to investigating what can be learnt from these competitions. Do they really generate scientific progress? What are common and successful participation strategies? What makes a solution superior to a competing method? To address this gap in the literature, we performed a multi-center study with all 80 competitions that were conducted in the scope of IEEE ISBI 2021 and MICCAI 2021. Statistical analyses performed based on comprehensive descriptions of the submitted algorithms linked to their rank as well as the underlying participation strategies revealed common characteristics of winning solutions. These typically include the use of multi-task learning (63%) and/or multi-stage pipelines (61%), and a focus on augmentation (100%), image preprocessing (97%), data curation (79%), and postprocessing (66%). The "typical" lead of a winning team is a computer scientist with a doctoral degree, five years of experience in biomedical image analysis, and four years of experience in deep learning. Two core general development strategies stood out for highly-ranked teams: the reflection of the metrics in the method design and the focus on analyzing and handling failure cases. According to the organizers, 43% of the winning algorithms exceeded the state of the art but only 11% completely solved the respective domain problem. The insights of our study could help researchers (1) improve algorithm development strategies when approaching new problems, and (2) focus on open research questions revealed by this work.
[ { "version": "v1", "created": "Thu, 30 Mar 2023 21:41:42 GMT" } ]
2023-04-03T00:00:00
[ [ "Eisenmann", "Matthias", "" ], [ "Reinke", "Annika", "" ], [ "Weru", "Vivienn", "" ], [ "Tizabi", "Minu Dietlinde", "" ], [ "Isensee", "Fabian", "" ], [ "Adler", "Tim J.", "" ], [ "Ali", "Sharib", "" ], [ "Andrearczyk", "Vincent", "" ], [ "Aubreville", "Marc", "" ], [ "Baid", "Ujjwal", "" ], [ "Bakas", "Spyridon", "" ], [ "Balu", "Niranjan", "" ], [ "Bano", "Sophia", "" ], [ "Bernal", "Jorge", "" ], [ "Bodenstedt", "Sebastian", "" ], [ "Casella", "Alessandro", "" ], [ "Cheplygina", "Veronika", "" ], [ "Daum", "Marie", "" ], [ "de Bruijne", "Marleen", "" ], [ "Depeursinge", "Adrien", "" ], [ "Dorent", "Reuben", "" ], [ "Egger", "Jan", "" ], [ "Ellis", "David G.", "" ], [ "Engelhardt", "Sandy", "" ], [ "Ganz", "Melanie", "" ], [ "Ghatwary", "Noha", "" ], [ "Girard", "Gabriel", "" ], [ "Godau", "Patrick", "" ], [ "Gupta", "Anubha", "" ], [ "Hansen", "Lasse", "" ], [ "Harada", "Kanako", "" ], [ "Heinrich", "Mattias", "" ], [ "Heller", "Nicholas", "" ], [ "Hering", "Alessa", "" ], [ "Huaulmé", "Arnaud", "" ], [ "Jannin", "Pierre", "" ], [ "Kavur", "Ali Emre", "" ], [ "Kodym", "Oldřich", "" ], [ "Kozubek", "Michal", "" ], [ "Li", "Jianning", "" ], [ "Li", "Hongwei", "" ], [ "Ma", "Jun", "" ], [ "Martín-Isla", "Carlos", "" ], [ "Menze", "Bjoern", "" ], [ "Noble", "Alison", "" ], [ "Oreiller", "Valentin", "" ], [ "Padoy", "Nicolas", "" ], [ "Pati", "Sarthak", "" ], [ "Payette", "Kelly", "" ], [ "Rädsch", "Tim", "" ], [ "Rafael-Patiño", "Jonathan", "" ], [ "Bawa", "Vivek Singh", "" ], [ "Speidel", "Stefanie", "" ], [ "Sudre", "Carole H.", "" ], [ "van Wijnen", "Kimberlin", "" ], [ "Wagner", "Martin", "" ], [ "Wei", "Donglai", "" ], [ "Yamlahi", "Amine", "" ], [ "Yap", "Moi Hoon", "" ], [ "Yuan", "Chun", "" ], [ "Zenk", "Maximilian", "" ], [ "Zia", "Aneeq", "" ], [ "Zimmerer", "David", "" ], [ "Aydogan", "Dogu Baran", "" ], [ "Bhattarai", "Binod", "" ], [ "Bloch", "Louise", "" ], [ "Brüngel", "Raphael", "" ], [ "Cho", "Jihoon", "" ], [ "Choi", "Chanyeol", "" ], [ "Dou", "Qi", "" ], [ "Ezhov", "Ivan", "" ], [ "Friedrich", "Christoph M.", "" ], [ "Fuller", "Clifton", "" ], [ "Gaire", "Rebati Raman", "" ], [ "Galdran", "Adrian", "" ], [ "Faura", "Álvaro García", "" ], [ "Grammatikopoulou", "Maria", "" ], [ "Hong", "SeulGi", "" ], [ "Jahanifar", "Mostafa", "" ], [ "Jang", "Ikbeom", "" ], [ "Kadkhodamohammadi", "Abdolrahim", "" ], [ "Kang", "Inha", "" ], [ "Kofler", "Florian", "" ], [ "Kondo", "Satoshi", "" ], [ "Kuijf", "Hugo", "" ], [ "Li", "Mingxing", "" ], [ "Luu", "Minh Huan", "" ], [ "Martinčič", "Tomaž", "" ], [ "Morais", "Pedro", "" ], [ "Naser", "Mohamed A.", "" ], [ "Oliveira", "Bruno", "" ], [ "Owen", "David", "" ], [ "Pang", "Subeen", "" ], [ "Park", "Jinah", "" ], [ "Park", "Sung-Hong", "" ], [ "Płotka", "Szymon", "" ], [ "Puybareau", "Elodie", "" ], [ "Rajpoot", "Nasir", "" ], [ "Ryu", "Kanghyun", "" ], [ "Saeed", "Numan", "" ], [ "Shephard", "Adam", "" ], [ "Shi", "Pengcheng", "" ], [ "Štepec", "Dejan", "" ], [ "Subedi", "Ronast", "" ], [ "Tochon", "Guillaume", "" ], [ "Torres", "Helena R.", "" ], [ "Urien", "Helene", "" ], [ "Vilaça", "João L.", "" ], [ "Wahid", "Kareem Abdul", "" ], [ "Wang", "Haojie", "" ], [ "Wang", "Jiacheng", "" ], [ "Wang", "Liansheng", "" ], [ "Wang", "Xiyue", "" ], [ "Wiestler", "Benedikt", "" ], [ "Wodzinski", "Marek", "" ], [ "Xia", "Fangfang", "" ], [ "Xie", "Juanying", "" ], [ "Xiong", "Zhiwei", "" ], [ "Yang", "Sen", "" ], [ "Yang", "Yanwu", "" ], [ "Zhao", "Zixuan", "" ], [ "Maier-Hein", "Klaus", "" ], [ "Jäger", "Paul F.", "" ], [ "Kopp-Schneider", "Annette", "" ], [ "Maier-Hein", "Lena", "" ] ]
TITLE: Why is the winner the best? ABSTRACT: International benchmarking competitions have become fundamental for the comparative performance assessment of image analysis methods. However, little attention has been given to investigating what can be learnt from these competitions. Do they really generate scientific progress? What are common and successful participation strategies? What makes a solution superior to a competing method? To address this gap in the literature, we performed a multi-center study with all 80 competitions that were conducted in the scope of IEEE ISBI 2021 and MICCAI 2021. Statistical analyses performed based on comprehensive descriptions of the submitted algorithms linked to their rank as well as the underlying participation strategies revealed common characteristics of winning solutions. These typically include the use of multi-task learning (63%) and/or multi-stage pipelines (61%), and a focus on augmentation (100%), image preprocessing (97%), data curation (79%), and postprocessing (66%). The "typical" lead of a winning team is a computer scientist with a doctoral degree, five years of experience in biomedical image analysis, and four years of experience in deep learning. Two core general development strategies stood out for highly-ranked teams: the reflection of the metrics in the method design and the focus on analyzing and handling failure cases. According to the organizers, 43% of the winning algorithms exceeded the state of the art but only 11% completely solved the respective domain problem. The insights of our study could help researchers (1) improve algorithm development strategies when approaching new problems, and (2) focus on open research questions revealed by this work.
no_new_dataset
0.707556
2303.17748
Mohammad Khodadad
Mohammad Khodadad, Morteza Rezanejad, Ali Shiraee Kasmaee, Kaleem Siddiqi, Dirk Walther, Hamidreza Mahyar
MLGCN: An Ultra Efficient Graph Convolution Neural Model For 3D Point Cloud Analysis
null
null
null
null
cs.CV
http://creativecommons.org/licenses/by/4.0/
The analysis of 3D point clouds has diverse applications in robotics, vision and graphics. Processing them presents specific challenges since they are naturally sparse, can vary in spatial resolution and are typically unordered. Graph-based networks to abstract features have emerged as a promising alternative to convolutional neural networks for their analysis, but these can be computationally heavy as well as memory inefficient. To address these limitations we introduce a novel Multi-level Graph Convolution Neural (MLGCN) model, which uses Graph Neural Networks (GNN) blocks to extract features from 3D point clouds at specific locality levels. Our approach employs precomputed graph KNNs, where each KNN graph is shared between GCN blocks inside a GNN block, making it both efficient and effective compared to present models. We demonstrate the efficacy of our approach on point cloud based object classification and part segmentation tasks on benchmark datasets, showing that it produces comparable results to those of state-of-the-art models while requiring up to a thousand times fewer floating-point operations (FLOPs) and having significantly reduced storage requirements. Thus, our MLGCN model could be particular relevant to point cloud based 3D shape analysis in industrial applications when computing resources are scarce.
[ { "version": "v1", "created": "Fri, 31 Mar 2023 00:15:22 GMT" } ]
2023-04-03T00:00:00
[ [ "Khodadad", "Mohammad", "" ], [ "Rezanejad", "Morteza", "" ], [ "Kasmaee", "Ali Shiraee", "" ], [ "Siddiqi", "Kaleem", "" ], [ "Walther", "Dirk", "" ], [ "Mahyar", "Hamidreza", "" ] ]
TITLE: MLGCN: An Ultra Efficient Graph Convolution Neural Model For 3D Point Cloud Analysis ABSTRACT: The analysis of 3D point clouds has diverse applications in robotics, vision and graphics. Processing them presents specific challenges since they are naturally sparse, can vary in spatial resolution and are typically unordered. Graph-based networks to abstract features have emerged as a promising alternative to convolutional neural networks for their analysis, but these can be computationally heavy as well as memory inefficient. To address these limitations we introduce a novel Multi-level Graph Convolution Neural (MLGCN) model, which uses Graph Neural Networks (GNN) blocks to extract features from 3D point clouds at specific locality levels. Our approach employs precomputed graph KNNs, where each KNN graph is shared between GCN blocks inside a GNN block, making it both efficient and effective compared to present models. We demonstrate the efficacy of our approach on point cloud based object classification and part segmentation tasks on benchmark datasets, showing that it produces comparable results to those of state-of-the-art models while requiring up to a thousand times fewer floating-point operations (FLOPs) and having significantly reduced storage requirements. Thus, our MLGCN model could be particular relevant to point cloud based 3D shape analysis in industrial applications when computing resources are scarce.
no_new_dataset
0.710672
2303.17766
Yongzhen Wang
Yongzhen Wang, Xuefeng Yan, Yanbiao Niu, Lina Gong, Yanwen Guo, Mingqiang Wei
Joint Depth Estimation and Mixture of Rain Removal From a Single Image
11 pages, 7 figures, 5 tables
null
null
null
cs.CV
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
Rainy weather significantly deteriorates the visibility of scene objects, particularly when images are captured through outdoor camera lenses or windshields. Through careful observation of numerous rainy photos, we have found that the images are generally affected by various rainwater artifacts such as raindrops, rain streaks, and rainy haze, which impact the image quality from both near and far distances, resulting in a complex and intertwined process of image degradation. However, current deraining techniques are limited in their ability to address only one or two types of rainwater, which poses a challenge in removing the mixture of rain (MOR). In this study, we propose an effective image deraining paradigm for Mixture of rain REmoval, called DEMore-Net, which takes full account of the MOR effect. Going beyond the existing deraining wisdom, DEMore-Net is a joint learning paradigm that integrates depth estimation and MOR removal tasks to achieve superior rain removal. The depth information can offer additional meaningful guidance information based on distance, thus better helping DEMore-Net remove different types of rainwater. Moreover, this study explores normalization approaches in image deraining tasks and introduces a new Hybrid Normalization Block (HNB) to enhance the deraining performance of DEMore-Net. Extensive experiments conducted on synthetic datasets and real-world MOR photos fully validate the superiority of the proposed DEMore-Net. Code is available at https://github.com/yz-wang/DEMore-Net.
[ { "version": "v1", "created": "Fri, 31 Mar 2023 02:05:45 GMT" } ]
2023-04-03T00:00:00
[ [ "Wang", "Yongzhen", "" ], [ "Yan", "Xuefeng", "" ], [ "Niu", "Yanbiao", "" ], [ "Gong", "Lina", "" ], [ "Guo", "Yanwen", "" ], [ "Wei", "Mingqiang", "" ] ]
TITLE: Joint Depth Estimation and Mixture of Rain Removal From a Single Image ABSTRACT: Rainy weather significantly deteriorates the visibility of scene objects, particularly when images are captured through outdoor camera lenses or windshields. Through careful observation of numerous rainy photos, we have found that the images are generally affected by various rainwater artifacts such as raindrops, rain streaks, and rainy haze, which impact the image quality from both near and far distances, resulting in a complex and intertwined process of image degradation. However, current deraining techniques are limited in their ability to address only one or two types of rainwater, which poses a challenge in removing the mixture of rain (MOR). In this study, we propose an effective image deraining paradigm for Mixture of rain REmoval, called DEMore-Net, which takes full account of the MOR effect. Going beyond the existing deraining wisdom, DEMore-Net is a joint learning paradigm that integrates depth estimation and MOR removal tasks to achieve superior rain removal. The depth information can offer additional meaningful guidance information based on distance, thus better helping DEMore-Net remove different types of rainwater. Moreover, this study explores normalization approaches in image deraining tasks and introduces a new Hybrid Normalization Block (HNB) to enhance the deraining performance of DEMore-Net. Extensive experiments conducted on synthetic datasets and real-world MOR photos fully validate the superiority of the proposed DEMore-Net. Code is available at https://github.com/yz-wang/DEMore-Net.
no_new_dataset
0.711418
2303.17769
Di Fan
Shaohan Chen, Di Fan and Chuanhou Gao
Domain Knowledge integrated for Blast Furnace Classifier Design
9 pages, 4 figures
null
null
null
cs.LG cs.AI
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
Blast furnace modeling and control is one of the important problems in the industrial field, and the black-box model is an effective mean to describe the complex blast furnace system. In practice, there are often different learning targets, such as safety and energy saving in industrial applications, depending on the application. For this reason, this paper proposes a framework to design a domain knowledge integrated classification model that yields a classifier for industrial application. Our knowledge incorporated learning scheme allows the users to create a classifier that identifies "important samples" (whose misclassifications can lead to severe consequences) more correctly, while keeping the proper precision of classifying the remaining samples. The effectiveness of the proposed method has been verified by two real blast furnace datasets, which guides the operators to utilize their prior experience for controlling the blast furnace systems better.
[ { "version": "v1", "created": "Fri, 31 Mar 2023 02:13:32 GMT" } ]
2023-04-03T00:00:00
[ [ "Chen", "Shaohan", "" ], [ "Fan", "Di", "" ], [ "Gao", "Chuanhou", "" ] ]
TITLE: Domain Knowledge integrated for Blast Furnace Classifier Design ABSTRACT: Blast furnace modeling and control is one of the important problems in the industrial field, and the black-box model is an effective mean to describe the complex blast furnace system. In practice, there are often different learning targets, such as safety and energy saving in industrial applications, depending on the application. For this reason, this paper proposes a framework to design a domain knowledge integrated classification model that yields a classifier for industrial application. Our knowledge incorporated learning scheme allows the users to create a classifier that identifies "important samples" (whose misclassifications can lead to severe consequences) more correctly, while keeping the proper precision of classifying the remaining samples. The effectiveness of the proposed method has been verified by two real blast furnace datasets, which guides the operators to utilize their prior experience for controlling the blast furnace systems better.
no_new_dataset
0.71403
2303.17782
Zann Koh
Zann Koh, Yan Qin, Yong Liang Guan, Chau Yuen
A Slow-Shifting Concerned Machine Learning Method for Short-term Traffic Flow Forecasting
6 pages, 4 figures. Accepted for IEEE International Conference on Smart Mobility 2023 (IEEE SM'23)
null
null
null
cs.LG
http://creativecommons.org/licenses/by-nc-nd/4.0/
The ability to predict traffic flow over time for crowded areas during rush hours is increasingly important as it can help authorities make informed decisions for congestion mitigation or scheduling of infrastructure development in an area. However, a crucial challenge in traffic flow forecasting is the slow shifting in temporal peaks between daily and weekly cycles, resulting in the nonstationarity of the traffic flow signal and leading to difficulty in accurate forecasting. To address this challenge, we propose a slow shifting concerned machine learning method for traffic flow forecasting, which includes two parts. First, we take advantage of Empirical Mode Decomposition as the feature engineering to alleviate the nonstationarity of traffic flow data, yielding a series of stationary components. Second, due to the superiority of Long-Short-Term-Memory networks in capturing temporal features, an advanced traffic flow forecasting model is developed by taking the stationary components as inputs. Finally, we apply this method on a benchmark of real-world data and provide a comparison with other existing methods. Our proposed method outperforms the state-of-art results by 14.55% and 62.56% using the metrics of root mean squared error and mean absolute percentage error, respectively.
[ { "version": "v1", "created": "Fri, 31 Mar 2023 03:07:53 GMT" } ]
2023-04-03T00:00:00
[ [ "Koh", "Zann", "" ], [ "Qin", "Yan", "" ], [ "Guan", "Yong Liang", "" ], [ "Yuen", "Chau", "" ] ]
TITLE: A Slow-Shifting Concerned Machine Learning Method for Short-term Traffic Flow Forecasting ABSTRACT: The ability to predict traffic flow over time for crowded areas during rush hours is increasingly important as it can help authorities make informed decisions for congestion mitigation or scheduling of infrastructure development in an area. However, a crucial challenge in traffic flow forecasting is the slow shifting in temporal peaks between daily and weekly cycles, resulting in the nonstationarity of the traffic flow signal and leading to difficulty in accurate forecasting. To address this challenge, we propose a slow shifting concerned machine learning method for traffic flow forecasting, which includes two parts. First, we take advantage of Empirical Mode Decomposition as the feature engineering to alleviate the nonstationarity of traffic flow data, yielding a series of stationary components. Second, due to the superiority of Long-Short-Term-Memory networks in capturing temporal features, an advanced traffic flow forecasting model is developed by taking the stationary components as inputs. Finally, we apply this method on a benchmark of real-world data and provide a comparison with other existing methods. Our proposed method outperforms the state-of-art results by 14.55% and 62.56% using the metrics of root mean squared error and mean absolute percentage error, respectively.
no_new_dataset
0.709648
2303.17799
Feng-Ju Chang
Feng-Ju Chang, Thejaswi Muniyappa, Kanthashree Mysore Sathyendra, Kai Wei, Grant P. Strimel, Ross McGowan
Dialog act guided contextual adapter for personalized speech recognition
Accepted at ICASSP 2023
null
null
null
cs.CL cs.SD eess.AS
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
Personalization in multi-turn dialogs has been a long standing challenge for end-to-end automatic speech recognition (E2E ASR) models. Recent work on contextual adapters has tackled rare word recognition using user catalogs. This adaptation, however, does not incorporate an important cue, the dialog act, which is available in a multi-turn dialog scenario. In this work, we propose a dialog act guided contextual adapter network. Specifically, it leverages dialog acts to select the most relevant user catalogs and creates queries based on both -- the audio as well as the semantic relationship between the carrier phrase and user catalogs to better guide the contextual biasing. On industrial voice assistant datasets, our model outperforms both the baselines - dialog act encoder-only model, and the contextual adaptation, leading to the most improvement over the no-context model: 58% average relative word error rate reduction (WERR) in the multi-turn dialog scenario, in comparison to the prior-art contextual adapter, which has achieved 39% WERR over the no-context model.
[ { "version": "v1", "created": "Fri, 31 Mar 2023 05:13:44 GMT" } ]
2023-04-03T00:00:00
[ [ "Chang", "Feng-Ju", "" ], [ "Muniyappa", "Thejaswi", "" ], [ "Sathyendra", "Kanthashree Mysore", "" ], [ "Wei", "Kai", "" ], [ "Strimel", "Grant P.", "" ], [ "McGowan", "Ross", "" ] ]
TITLE: Dialog act guided contextual adapter for personalized speech recognition ABSTRACT: Personalization in multi-turn dialogs has been a long standing challenge for end-to-end automatic speech recognition (E2E ASR) models. Recent work on contextual adapters has tackled rare word recognition using user catalogs. This adaptation, however, does not incorporate an important cue, the dialog act, which is available in a multi-turn dialog scenario. In this work, we propose a dialog act guided contextual adapter network. Specifically, it leverages dialog acts to select the most relevant user catalogs and creates queries based on both -- the audio as well as the semantic relationship between the carrier phrase and user catalogs to better guide the contextual biasing. On industrial voice assistant datasets, our model outperforms both the baselines - dialog act encoder-only model, and the contextual adaptation, leading to the most improvement over the no-context model: 58% average relative word error rate reduction (WERR) in the multi-turn dialog scenario, in comparison to the prior-art contextual adapter, which has achieved 39% WERR over the no-context model.
no_new_dataset
0.710666
2303.17815
Hengjia Li
Hengjia Li, Tu Zheng, Zhihao Chi, Zheng Yang, Wenxiao Wang, Boxi Wu, Binbin Lin, Deng Cai
APPT : Asymmetric Parallel Point Transformer for 3D Point Cloud Understanding
null
null
null
null
cs.CV
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
Transformer-based networks have achieved impressive performance in 3D point cloud understanding. However, most of them concentrate on aggregating local features, but neglect to directly model global dependencies, which results in a limited effective receptive field. Besides, how to effectively incorporate local and global components also remains challenging. To tackle these problems, we propose Asymmetric Parallel Point Transformer (APPT). Specifically, we introduce Global Pivot Attention to extract global features and enlarge the effective receptive field. Moreover, we design the Asymmetric Parallel structure to effectively integrate local and global information. Combined with these designs, APPT is able to capture features globally throughout the entire network while focusing on local-detailed features. Extensive experiments show that our method outperforms the priors and achieves state-of-the-art on several benchmarks for 3D point cloud understanding, such as 3D semantic segmentation on S3DIS, 3D shape classification on ModelNet40, and 3D part segmentation on ShapeNet.
[ { "version": "v1", "created": "Fri, 31 Mar 2023 06:11:02 GMT" } ]
2023-04-03T00:00:00
[ [ "Li", "Hengjia", "" ], [ "Zheng", "Tu", "" ], [ "Chi", "Zhihao", "" ], [ "Yang", "Zheng", "" ], [ "Wang", "Wenxiao", "" ], [ "Wu", "Boxi", "" ], [ "Lin", "Binbin", "" ], [ "Cai", "Deng", "" ] ]
TITLE: APPT : Asymmetric Parallel Point Transformer for 3D Point Cloud Understanding ABSTRACT: Transformer-based networks have achieved impressive performance in 3D point cloud understanding. However, most of them concentrate on aggregating local features, but neglect to directly model global dependencies, which results in a limited effective receptive field. Besides, how to effectively incorporate local and global components also remains challenging. To tackle these problems, we propose Asymmetric Parallel Point Transformer (APPT). Specifically, we introduce Global Pivot Attention to extract global features and enlarge the effective receptive field. Moreover, we design the Asymmetric Parallel structure to effectively integrate local and global information. Combined with these designs, APPT is able to capture features globally throughout the entire network while focusing on local-detailed features. Extensive experiments show that our method outperforms the priors and achieves state-of-the-art on several benchmarks for 3D point cloud understanding, such as 3D semantic segmentation on S3DIS, 3D shape classification on ModelNet40, and 3D part segmentation on ShapeNet.
no_new_dataset
0.710603
2303.17820
Xiaoyu Zhang
Xiaoyu Zhang, Xiwei Xuan, Alden Dima, Thurston Sexton, Kwan-Liu Ma
LabelVizier: Interactive Validation and Relabeling for Technical Text Annotations
10 pages, 5 figures
null
null
null
cs.HC
http://creativecommons.org/licenses/by-nc-nd/4.0/
With the rapid accumulation of text data produced by data-driven techniques, the task of extracting "data annotations"--concise, high-quality data summaries from unstructured raw text--has become increasingly important. The recent advances in weak supervision and crowd-sourcing techniques provide promising solutions to efficiently create annotations (labels) for large-scale technical text data. However, such annotations may fail in practice because of the change in annotation requirements, application scenarios, and modeling goals, where label validation and relabeling by domain experts are required. To approach this issue, we present LabelVizier, a human-in-the-loop workflow that incorporates domain knowledge and user-specific requirements to reveal actionable insights into annotation flaws, then produce better-quality labels for large-scale multi-label datasets. We implement our workflow as an interactive notebook to facilitate flexible error profiling, in-depth annotation validation for three error types, and efficient annotation relabeling on different data scales. We evaluated the efficiency and generalizability of our workflow with two use cases and four expert reviews. The results indicate that LabelVizier is applicable in various application scenarios and assist domain experts with different knowledge backgrounds to efficiently improve technical text annotation quality.
[ { "version": "v1", "created": "Fri, 31 Mar 2023 06:31:06 GMT" } ]
2023-04-03T00:00:00
[ [ "Zhang", "Xiaoyu", "" ], [ "Xuan", "Xiwei", "" ], [ "Dima", "Alden", "" ], [ "Sexton", "Thurston", "" ], [ "Ma", "Kwan-Liu", "" ] ]
TITLE: LabelVizier: Interactive Validation and Relabeling for Technical Text Annotations ABSTRACT: With the rapid accumulation of text data produced by data-driven techniques, the task of extracting "data annotations"--concise, high-quality data summaries from unstructured raw text--has become increasingly important. The recent advances in weak supervision and crowd-sourcing techniques provide promising solutions to efficiently create annotations (labels) for large-scale technical text data. However, such annotations may fail in practice because of the change in annotation requirements, application scenarios, and modeling goals, where label validation and relabeling by domain experts are required. To approach this issue, we present LabelVizier, a human-in-the-loop workflow that incorporates domain knowledge and user-specific requirements to reveal actionable insights into annotation flaws, then produce better-quality labels for large-scale multi-label datasets. We implement our workflow as an interactive notebook to facilitate flexible error profiling, in-depth annotation validation for three error types, and efficient annotation relabeling on different data scales. We evaluated the efficiency and generalizability of our workflow with two use cases and four expert reviews. The results indicate that LabelVizier is applicable in various application scenarios and assist domain experts with different knowledge backgrounds to efficiently improve technical text annotation quality.
no_new_dataset
0.708603
2303.17836
Naveed Akhtar Dr.
Naveed Akhtar, Mohammad A. A. K. Jalwana
Rethinking interpretation: Input-agnostic saliency mapping of deep visual classifiers
Accepted for publication in AAAI 2023
null
null
null
cs.CV cs.AI cs.LG
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
Saliency methods provide post-hoc model interpretation by attributing input features to the model outputs. Current methods mainly achieve this using a single input sample, thereby failing to answer input-independent inquiries about the model. We also show that input-specific saliency mapping is intrinsically susceptible to misleading feature attribution. Current attempts to use 'general' input features for model interpretation assume access to a dataset containing those features, which biases the interpretation. Addressing the gap, we introduce a new perspective of input-agnostic saliency mapping that computationally estimates the high-level features attributed by the model to its outputs. These features are geometrically correlated, and are computed by accumulating model's gradient information with respect to an unrestricted data distribution. To compute these features, we nudge independent data points over the model loss surface towards the local minima associated by a human-understandable concept, e.g., class label for classifiers. With a systematic projection, scaling and refinement process, this information is transformed into an interpretable visualization without compromising its model-fidelity. The visualization serves as a stand-alone qualitative interpretation. With an extensive evaluation, we not only demonstrate successful visualizations for a variety of concepts for large-scale models, but also showcase an interesting utility of this new form of saliency mapping by identifying backdoor signatures in compromised classifiers.
[ { "version": "v1", "created": "Fri, 31 Mar 2023 06:58:45 GMT" } ]
2023-04-03T00:00:00
[ [ "Akhtar", "Naveed", "" ], [ "Jalwana", "Mohammad A. A. K.", "" ] ]
TITLE: Rethinking interpretation: Input-agnostic saliency mapping of deep visual classifiers ABSTRACT: Saliency methods provide post-hoc model interpretation by attributing input features to the model outputs. Current methods mainly achieve this using a single input sample, thereby failing to answer input-independent inquiries about the model. We also show that input-specific saliency mapping is intrinsically susceptible to misleading feature attribution. Current attempts to use 'general' input features for model interpretation assume access to a dataset containing those features, which biases the interpretation. Addressing the gap, we introduce a new perspective of input-agnostic saliency mapping that computationally estimates the high-level features attributed by the model to its outputs. These features are geometrically correlated, and are computed by accumulating model's gradient information with respect to an unrestricted data distribution. To compute these features, we nudge independent data points over the model loss surface towards the local minima associated by a human-understandable concept, e.g., class label for classifiers. With a systematic projection, scaling and refinement process, this information is transformed into an interpretable visualization without compromising its model-fidelity. The visualization serves as a stand-alone qualitative interpretation. With an extensive evaluation, we not only demonstrate successful visualizations for a variety of concepts for large-scale models, but also showcase an interesting utility of this new form of saliency mapping by identifying backdoor signatures in compromised classifiers.
no_new_dataset
0.710226
2303.17839
Yiwu Zhong
Yiwu Zhong, Licheng Yu, Yang Bai, Shangwen Li, Xueting Yan, Yin Li
Learning Procedure-aware Video Representation from Instructional Videos and Their Narrations
Accepted to CVPR 2023
null
null
null
cs.CV cs.AI cs.CL cs.LG cs.MM
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
The abundance of instructional videos and their narrations over the Internet offers an exciting avenue for understanding procedural activities. In this work, we propose to learn video representation that encodes both action steps and their temporal ordering, based on a large-scale dataset of web instructional videos and their narrations, without using human annotations. Our method jointly learns a video representation to encode individual step concepts, and a deep probabilistic model to capture both temporal dependencies and immense individual variations in the step ordering. We empirically demonstrate that learning temporal ordering not only enables new capabilities for procedure reasoning, but also reinforces the recognition of individual steps. Our model significantly advances the state-of-the-art results on step classification (+2.8% / +3.3% on COIN / EPIC-Kitchens) and step forecasting (+7.4% on COIN). Moreover, our model attains promising results in zero-shot inference for step classification and forecasting, as well as in predicting diverse and plausible steps for incomplete procedures. Our code is available at https://github.com/facebookresearch/ProcedureVRL.
[ { "version": "v1", "created": "Fri, 31 Mar 2023 07:02:26 GMT" } ]
2023-04-03T00:00:00
[ [ "Zhong", "Yiwu", "" ], [ "Yu", "Licheng", "" ], [ "Bai", "Yang", "" ], [ "Li", "Shangwen", "" ], [ "Yan", "Xueting", "" ], [ "Li", "Yin", "" ] ]
TITLE: Learning Procedure-aware Video Representation from Instructional Videos and Their Narrations ABSTRACT: The abundance of instructional videos and their narrations over the Internet offers an exciting avenue for understanding procedural activities. In this work, we propose to learn video representation that encodes both action steps and their temporal ordering, based on a large-scale dataset of web instructional videos and their narrations, without using human annotations. Our method jointly learns a video representation to encode individual step concepts, and a deep probabilistic model to capture both temporal dependencies and immense individual variations in the step ordering. We empirically demonstrate that learning temporal ordering not only enables new capabilities for procedure reasoning, but also reinforces the recognition of individual steps. Our model significantly advances the state-of-the-art results on step classification (+2.8% / +3.3% on COIN / EPIC-Kitchens) and step forecasting (+7.4% on COIN). Moreover, our model attains promising results in zero-shot inference for step classification and forecasting, as well as in predicting diverse and plausible steps for incomplete procedures. Our code is available at https://github.com/facebookresearch/ProcedureVRL.
no_new_dataset
0.71086
2303.17842
Jinwoo Kim
Jinwoo Kim, Janghyuk Choi, Ho-Jin Choi, Seon Joo Kim
Shepherding Slots to Objects: Towards Stable and Robust Object-Centric Learning
null
null
null
null
cs.CV
http://creativecommons.org/licenses/by/4.0/
Object-centric learning (OCL) aspires general and compositional understanding of scenes by representing a scene as a collection of object-centric representations. OCL has also been extended to multi-view image and video datasets to apply various data-driven inductive biases by utilizing geometric or temporal information in the multi-image data. Single-view images carry less information about how to disentangle a given scene than videos or multi-view images do. Hence, owing to the difficulty of applying inductive biases, OCL for single-view images remains challenging, resulting in inconsistent learning of object-centric representation. To this end, we introduce a novel OCL framework for single-view images, SLot Attention via SHepherding (SLASH), which consists of two simple-yet-effective modules on top of Slot Attention. The new modules, Attention Refining Kernel (ARK) and Intermediate Point Predictor and Encoder (IPPE), respectively, prevent slots from being distracted by the background noise and indicate locations for slots to focus on to facilitate learning of object-centric representation. We also propose a weak semi-supervision approach for OCL, whilst our proposed framework can be used without any assistant annotation during the inference. Experiments show that our proposed method enables consistent learning of object-centric representation and achieves strong performance across four datasets. Code is available at \url{https://github.com/object-understanding/SLASH}.
[ { "version": "v1", "created": "Fri, 31 Mar 2023 07:07:29 GMT" } ]
2023-04-03T00:00:00
[ [ "Kim", "Jinwoo", "" ], [ "Choi", "Janghyuk", "" ], [ "Choi", "Ho-Jin", "" ], [ "Kim", "Seon Joo", "" ] ]
TITLE: Shepherding Slots to Objects: Towards Stable and Robust Object-Centric Learning ABSTRACT: Object-centric learning (OCL) aspires general and compositional understanding of scenes by representing a scene as a collection of object-centric representations. OCL has also been extended to multi-view image and video datasets to apply various data-driven inductive biases by utilizing geometric or temporal information in the multi-image data. Single-view images carry less information about how to disentangle a given scene than videos or multi-view images do. Hence, owing to the difficulty of applying inductive biases, OCL for single-view images remains challenging, resulting in inconsistent learning of object-centric representation. To this end, we introduce a novel OCL framework for single-view images, SLot Attention via SHepherding (SLASH), which consists of two simple-yet-effective modules on top of Slot Attention. The new modules, Attention Refining Kernel (ARK) and Intermediate Point Predictor and Encoder (IPPE), respectively, prevent slots from being distracted by the background noise and indicate locations for slots to focus on to facilitate learning of object-centric representation. We also propose a weak semi-supervision approach for OCL, whilst our proposed framework can be used without any assistant annotation during the inference. Experiments show that our proposed method enables consistent learning of object-centric representation and achieves strong performance across four datasets. Code is available at \url{https://github.com/object-understanding/SLASH}.
no_new_dataset
0.709629
2303.17845
Ayokunle Ige
Ayokunle Olalekan Ige, Mohd Halim Mohd Noor
WSense: A Robust Feature Learning Module for Lightweight Human Activity Recognition
null
null
null
null
cs.CV
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
In recent times, various modules such as squeeze-and-excitation, and others have been proposed to improve the quality of features learned from wearable sensor signals. However, these modules often cause the number of parameters to be large, which is not suitable for building lightweight human activity recognition models which can be easily deployed on end devices. In this research, we propose a feature learning module, termed WSense, which uses two 1D CNN and global max pooling layers to extract similar quality features from wearable sensor data while ignoring the difference in activity recognition models caused by the size of the sliding window. Experiments were carried out using CNN and ConvLSTM feature learning pipelines on a dataset obtained with a single accelerometer (WISDM) and another obtained using the fusion of accelerometers, gyroscopes, and magnetometers (PAMAP2) under various sliding window sizes. A total of nine hundred sixty (960) experiments were conducted to validate the WSense module against baselines and existing methods on the two datasets. The results showed that the WSense module aided pipelines in learning similar quality features and outperformed the baselines and existing models with a minimal and uniform model size across all sliding window segmentations. The code is available at https://github.com/AOige/WSense.
[ { "version": "v1", "created": "Fri, 31 Mar 2023 07:12:58 GMT" } ]
2023-04-03T00:00:00
[ [ "Ige", "Ayokunle Olalekan", "" ], [ "Noor", "Mohd Halim Mohd", "" ] ]
TITLE: WSense: A Robust Feature Learning Module for Lightweight Human Activity Recognition ABSTRACT: In recent times, various modules such as squeeze-and-excitation, and others have been proposed to improve the quality of features learned from wearable sensor signals. However, these modules often cause the number of parameters to be large, which is not suitable for building lightweight human activity recognition models which can be easily deployed on end devices. In this research, we propose a feature learning module, termed WSense, which uses two 1D CNN and global max pooling layers to extract similar quality features from wearable sensor data while ignoring the difference in activity recognition models caused by the size of the sliding window. Experiments were carried out using CNN and ConvLSTM feature learning pipelines on a dataset obtained with a single accelerometer (WISDM) and another obtained using the fusion of accelerometers, gyroscopes, and magnetometers (PAMAP2) under various sliding window sizes. A total of nine hundred sixty (960) experiments were conducted to validate the WSense module against baselines and existing methods on the two datasets. The results showed that the WSense module aided pipelines in learning similar quality features and outperformed the baselines and existing models with a minimal and uniform model size across all sliding window segmentations. The code is available at https://github.com/AOige/WSense.
no_new_dataset
0.711425
2303.17849
Takayuki Miura
Takayuki Miura, Toshiki Shibahara, Masanobu Kii, Atsunori Ichikawa, Juko Yamamoto, Koji Chida
On R\'{e}nyi Differential Privacy in Statistics-Based Synthetic Data Generation
18 pages, 3 figures
null
null
null
cs.CR cs.IT math.IT
http://creativecommons.org/licenses/by/4.0/
Privacy protection with synthetic data generation often uses differentially private statistics and model parameters to quantitatively express theoretical security. However, these methods do not take into account privacy protection due to the randomness of data generation. In this paper, we theoretically evaluate R\'{e}nyi differential privacy of the randomness in data generation of a synthetic data generation method that uses the mean vector and the covariance matrix of an original dataset. Specifically, for a fixed $\alpha > 1$, we show the condition of $\varepsilon$ such that the synthetic data generation satisfies $(\alpha, \varepsilon)$-R\'{e}nyi differential privacy under a bounded neighboring condition and an unbounded neighboring condition, respectively. In particular, under the unbounded condition, when the size of the original dataset and synthetic datase is 10 million, the mechanism satisfies $(4, 0.576)$-R\'{e}nyi differential privacy. We also show that when we translate it into the traditional $(\varepsilon, \delta)$-differential privacy, the mechanism satisfies $(4.00, 10^{-10})$-differential privacy.
[ { "version": "v1", "created": "Fri, 31 Mar 2023 07:26:52 GMT" } ]
2023-04-03T00:00:00
[ [ "Miura", "Takayuki", "" ], [ "Shibahara", "Toshiki", "" ], [ "Kii", "Masanobu", "" ], [ "Ichikawa", "Atsunori", "" ], [ "Yamamoto", "Juko", "" ], [ "Chida", "Koji", "" ] ]
TITLE: On R\'{e}nyi Differential Privacy in Statistics-Based Synthetic Data Generation ABSTRACT: Privacy protection with synthetic data generation often uses differentially private statistics and model parameters to quantitatively express theoretical security. However, these methods do not take into account privacy protection due to the randomness of data generation. In this paper, we theoretically evaluate R\'{e}nyi differential privacy of the randomness in data generation of a synthetic data generation method that uses the mean vector and the covariance matrix of an original dataset. Specifically, for a fixed $\alpha > 1$, we show the condition of $\varepsilon$ such that the synthetic data generation satisfies $(\alpha, \varepsilon)$-R\'{e}nyi differential privacy under a bounded neighboring condition and an unbounded neighboring condition, respectively. In particular, under the unbounded condition, when the size of the original dataset and synthetic datase is 10 million, the mechanism satisfies $(4, 0.576)$-R\'{e}nyi differential privacy. We also show that when we translate it into the traditional $(\varepsilon, \delta)$-differential privacy, the mechanism satisfies $(4.00, 10^{-10})$-differential privacy.
no_new_dataset
0.710986
2303.17852
Qian Wang
Qian Wang, Kamran Paynabar
Maximum Covariance Unfolding Regression: A Novel Covariate-based Manifold Learning Approach for Point Cloud Data
null
null
null
null
cs.LG cs.CV
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
Point cloud data are widely used in manufacturing applications for process inspection, modeling, monitoring and optimization. The state-of-art tensor regression techniques have effectively been used for analysis of structured point cloud data, where the measurements on a uniform grid can be formed into a tensor. However, these techniques are not capable of handling unstructured point cloud data that are often in the form of manifolds. In this paper, we propose a nonlinear dimension reduction approach named Maximum Covariance Unfolding Regression that is able to learn the low-dimensional (LD) manifold of point clouds with the highest correlation with explanatory covariates. This LD manifold is then used for regression modeling and process optimization based on process variables. The performance of the proposed method is subsequently evaluated and compared with benchmark methods through simulations and a case study of steel bracket manufacturing.
[ { "version": "v1", "created": "Fri, 31 Mar 2023 07:29:36 GMT" } ]
2023-04-03T00:00:00
[ [ "Wang", "Qian", "" ], [ "Paynabar", "Kamran", "" ] ]
TITLE: Maximum Covariance Unfolding Regression: A Novel Covariate-based Manifold Learning Approach for Point Cloud Data ABSTRACT: Point cloud data are widely used in manufacturing applications for process inspection, modeling, monitoring and optimization. The state-of-art tensor regression techniques have effectively been used for analysis of structured point cloud data, where the measurements on a uniform grid can be formed into a tensor. However, these techniques are not capable of handling unstructured point cloud data that are often in the form of manifolds. In this paper, we propose a nonlinear dimension reduction approach named Maximum Covariance Unfolding Regression that is able to learn the low-dimensional (LD) manifold of point clouds with the highest correlation with explanatory covariates. This LD manifold is then used for regression modeling and process optimization based on process variables. The performance of the proposed method is subsequently evaluated and compared with benchmark methods through simulations and a case study of steel bracket manufacturing.
no_new_dataset
0.710189
2303.17882
Haiming Yao
Haiming Yao, Wei Luo, Wenyong Yu
Visual Anomaly Detection via Dual-Attention Transformer and Discriminative Flow
Submission to IEEE Transactions On Industrial Informatics
null
null
null
cs.CV
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
In this paper, we introduce the novel state-of-the-art Dual-attention Transformer and Discriminative Flow (DADF) framework for visual anomaly detection. Based on only normal knowledge, visual anomaly detection has wide applications in industrial scenarios and has attracted significant attention. However, most existing methods fail to meet the requirements. In contrast, the proposed DTDF presents a new paradigm: it firstly leverages a pre-trained network to acquire multi-scale prior embeddings, followed by the development of a vision Transformer with dual attention mechanisms, namely self-attention and memorial-attention, to achieve two-level reconstruction for prior embeddings with the sequential and normality association. Additionally, we propose using normalizing flow to establish discriminative likelihood for the joint distribution of prior and reconstructions at each scale. The DADF achieves 98.3/98.4 of image/pixel AUROC on Mvtec AD; 83.7 of image AUROC and 67.4 of pixel sPRO on Mvtec LOCO AD benchmarks, demonstrating the effectiveness of our proposed approach.
[ { "version": "v1", "created": "Fri, 31 Mar 2023 08:34:07 GMT" } ]
2023-04-03T00:00:00
[ [ "Yao", "Haiming", "" ], [ "Luo", "Wei", "" ], [ "Yu", "Wenyong", "" ] ]
TITLE: Visual Anomaly Detection via Dual-Attention Transformer and Discriminative Flow ABSTRACT: In this paper, we introduce the novel state-of-the-art Dual-attention Transformer and Discriminative Flow (DADF) framework for visual anomaly detection. Based on only normal knowledge, visual anomaly detection has wide applications in industrial scenarios and has attracted significant attention. However, most existing methods fail to meet the requirements. In contrast, the proposed DTDF presents a new paradigm: it firstly leverages a pre-trained network to acquire multi-scale prior embeddings, followed by the development of a vision Transformer with dual attention mechanisms, namely self-attention and memorial-attention, to achieve two-level reconstruction for prior embeddings with the sequential and normality association. Additionally, we propose using normalizing flow to establish discriminative likelihood for the joint distribution of prior and reconstructions at each scale. The DADF achieves 98.3/98.4 of image/pixel AUROC on Mvtec AD; 83.7 of image AUROC and 67.4 of pixel sPRO on Mvtec LOCO AD benchmarks, demonstrating the effectiveness of our proposed approach.
no_new_dataset
0.707992
2303.17885
Yanjie Dong
Yanjie Dong, Luya Wang, Yuanfang Chi, Jia Wang, Haijun Zhang, Fei Richard Yu, Victor C. M. Leung, Xiping Hu
Accelerating Wireless Federated Learning via Nesterov's Momentum and Distributed Principle Component Analysis
null
null
null
null
cs.LG cs.IT math.IT
http://creativecommons.org/licenses/by/4.0/
A wireless federated learning system is investigated by allowing a server and workers to exchange uncoded information via orthogonal wireless channels. Since the workers frequently upload local gradients to the server via bandwidth-limited channels, the uplink transmission from the workers to the server becomes a communication bottleneck. Therefore, a one-shot distributed principle component analysis (PCA) is leveraged to reduce the dimension of uploaded gradients such that the communication bottleneck is relieved. A PCA-based wireless federated learning (PCA-WFL) algorithm and its accelerated version (i.e., PCA-AWFL) are proposed based on the low-dimensional gradients and the Nesterov's momentum. For the non-convex loss functions, a finite-time analysis is performed to quantify the impacts of system hyper-parameters on the convergence of the PCA-WFL and PCA-AWFL algorithms. The PCA-AWFL algorithm is theoretically certified to converge faster than the PCA-WFL algorithm. Besides, the convergence rates of PCA-WFL and PCA-AWFL algorithms quantitatively reveal the linear speedup with respect to the number of workers over the vanilla gradient descent algorithm. Numerical results are used to demonstrate the improved convergence rates of the proposed PCA-WFL and PCA-AWFL algorithms over the benchmarks.
[ { "version": "v1", "created": "Fri, 31 Mar 2023 08:41:42 GMT" } ]
2023-04-03T00:00:00
[ [ "Dong", "Yanjie", "" ], [ "Wang", "Luya", "" ], [ "Chi", "Yuanfang", "" ], [ "Wang", "Jia", "" ], [ "Zhang", "Haijun", "" ], [ "Yu", "Fei Richard", "" ], [ "Leung", "Victor C. M.", "" ], [ "Hu", "Xiping", "" ] ]
TITLE: Accelerating Wireless Federated Learning via Nesterov's Momentum and Distributed Principle Component Analysis ABSTRACT: A wireless federated learning system is investigated by allowing a server and workers to exchange uncoded information via orthogonal wireless channels. Since the workers frequently upload local gradients to the server via bandwidth-limited channels, the uplink transmission from the workers to the server becomes a communication bottleneck. Therefore, a one-shot distributed principle component analysis (PCA) is leveraged to reduce the dimension of uploaded gradients such that the communication bottleneck is relieved. A PCA-based wireless federated learning (PCA-WFL) algorithm and its accelerated version (i.e., PCA-AWFL) are proposed based on the low-dimensional gradients and the Nesterov's momentum. For the non-convex loss functions, a finite-time analysis is performed to quantify the impacts of system hyper-parameters on the convergence of the PCA-WFL and PCA-AWFL algorithms. The PCA-AWFL algorithm is theoretically certified to converge faster than the PCA-WFL algorithm. Besides, the convergence rates of PCA-WFL and PCA-AWFL algorithms quantitatively reveal the linear speedup with respect to the number of workers over the vanilla gradient descent algorithm. Numerical results are used to demonstrate the improved convergence rates of the proposed PCA-WFL and PCA-AWFL algorithms over the benchmarks.
no_new_dataset
0.71039
2303.17905
Jianfeng Xiang
Jianfeng Xiang, Jiaolong Yang, Binbin Huang, Xin Tong
3D-aware Image Generation using 2D Diffusion Models
Website: https://jeffreyxiang.github.io/ivid/
null
null
null
cs.CV
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
In this paper, we introduce a novel 3D-aware image generation method that leverages 2D diffusion models. We formulate the 3D-aware image generation task as multiview 2D image set generation, and further to a sequential unconditional-conditional multiview image generation process. This allows us to utilize 2D diffusion models to boost the generative modeling power of the method. Additionally, we incorporate depth information from monocular depth estimators to construct the training data for the conditional diffusion model using only still images. We train our method on a large-scale dataset, i.e., ImageNet, which is not addressed by previous methods. It produces high-quality images that significantly outperform prior methods. Furthermore, our approach showcases its capability to generate instances with large view angles, even though the training images are diverse and unaligned, gathered from "in-the-wild" real-world environments.
[ { "version": "v1", "created": "Fri, 31 Mar 2023 09:03:18 GMT" } ]
2023-04-03T00:00:00
[ [ "Xiang", "Jianfeng", "" ], [ "Yang", "Jiaolong", "" ], [ "Huang", "Binbin", "" ], [ "Tong", "Xin", "" ] ]
TITLE: 3D-aware Image Generation using 2D Diffusion Models ABSTRACT: In this paper, we introduce a novel 3D-aware image generation method that leverages 2D diffusion models. We formulate the 3D-aware image generation task as multiview 2D image set generation, and further to a sequential unconditional-conditional multiview image generation process. This allows us to utilize 2D diffusion models to boost the generative modeling power of the method. Additionally, we incorporate depth information from monocular depth estimators to construct the training data for the conditional diffusion model using only still images. We train our method on a large-scale dataset, i.e., ImageNet, which is not addressed by previous methods. It produces high-quality images that significantly outperform prior methods. Furthermore, our approach showcases its capability to generate instances with large view angles, even though the training images are diverse and unaligned, gathered from "in-the-wild" real-world environments.
no_new_dataset
0.71302
2303.17912
Jo\~ao Pedro Ara\'ujo
Joao Pedro Araujo, Jiaman Li, Karthik Vetrivel, Rishi Agarwal, Deepak Gopinath, Jiajun Wu, Alexander Clegg, C. Karen Liu
CIRCLE: Capture In Rich Contextual Environments
null
null
null
null
cs.CV cs.GR
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
Synthesizing 3D human motion in a contextual, ecological environment is important for simulating realistic activities people perform in the real world. However, conventional optics-based motion capture systems are not suited for simultaneously capturing human movements and complex scenes. The lack of rich contextual 3D human motion datasets presents a roadblock to creating high-quality generative human motion models. We propose a novel motion acquisition system in which the actor perceives and operates in a highly contextual virtual world while being motion captured in the real world. Our system enables rapid collection of high-quality human motion in highly diverse scenes, without the concern of occlusion or the need for physical scene construction in the real world. We present CIRCLE, a dataset containing 10 hours of full-body reaching motion from 5 subjects across nine scenes, paired with ego-centric information of the environment represented in various forms, such as RGBD videos. We use this dataset to train a model that generates human motion conditioned on scene information. Leveraging our dataset, the model learns to use ego-centric scene information to achieve nontrivial reaching tasks in the context of complex 3D scenes. To download the data please visit https://stanford-tml.github.io/circle_dataset/.
[ { "version": "v1", "created": "Fri, 31 Mar 2023 09:18:12 GMT" } ]
2023-04-03T00:00:00
[ [ "Araujo", "Joao Pedro", "" ], [ "Li", "Jiaman", "" ], [ "Vetrivel", "Karthik", "" ], [ "Agarwal", "Rishi", "" ], [ "Gopinath", "Deepak", "" ], [ "Wu", "Jiajun", "" ], [ "Clegg", "Alexander", "" ], [ "Liu", "C. Karen", "" ] ]
TITLE: CIRCLE: Capture In Rich Contextual Environments ABSTRACT: Synthesizing 3D human motion in a contextual, ecological environment is important for simulating realistic activities people perform in the real world. However, conventional optics-based motion capture systems are not suited for simultaneously capturing human movements and complex scenes. The lack of rich contextual 3D human motion datasets presents a roadblock to creating high-quality generative human motion models. We propose a novel motion acquisition system in which the actor perceives and operates in a highly contextual virtual world while being motion captured in the real world. Our system enables rapid collection of high-quality human motion in highly diverse scenes, without the concern of occlusion or the need for physical scene construction in the real world. We present CIRCLE, a dataset containing 10 hours of full-body reaching motion from 5 subjects across nine scenes, paired with ego-centric information of the environment represented in various forms, such as RGBD videos. We use this dataset to train a model that generates human motion conditioned on scene information. Leveraging our dataset, the model learns to use ego-centric scene information to achieve nontrivial reaching tasks in the context of complex 3D scenes. To download the data please visit https://stanford-tml.github.io/circle_dataset/.
new_dataset
0.715989
2303.17915
Debayan Bhattacharya
Debayan Bhattacharya, Finn Behrendt, Benjamin Tobias Becker, Dirk Beyersdorff, Elina Petersen, Marvin Petersen, Bastian Cheng, Dennis Eggert, Christian Betz, Anna Sophie Hoffmann, Alexander Schlaefer
Multiple Instance Ensembling For Paranasal Anomaly Classification In The Maxillary Sinus
null
null
null
null
eess.IV cs.CV
http://creativecommons.org/licenses/by/4.0/
Paranasal anomalies are commonly discovered during routine radiological screenings and can present with a wide range of morphological features. This diversity can make it difficult for convolutional neural networks (CNNs) to accurately classify these anomalies, especially when working with limited datasets. Additionally, current approaches to paranasal anomaly classification are constrained to identifying a single anomaly at a time. These challenges necessitate the need for further research and development in this area. In this study, we investigate the feasibility of using a 3D convolutional neural network (CNN) to classify healthy maxillary sinuses (MS) and MS with polyps or cysts. The task of accurately identifying the relevant MS volume within larger head and neck Magnetic Resonance Imaging (MRI) scans can be difficult, but we develop a straightforward strategy to tackle this challenge. Our end-to-end solution includes the use of a novel sampling technique that not only effectively localizes the relevant MS volume, but also increases the size of the training dataset and improves classification results. Additionally, we employ a multiple instance ensemble prediction method to further boost classification performance. Finally, we identify the optimal size of MS volumes to achieve the highest possible classification performance on our dataset. With our multiple instance ensemble prediction strategy and sampling strategy, our 3D CNNs achieve an F1 of 0.85 whereas without it, they achieve an F1 of 0.70. We demonstrate the feasibility of classifying anomalies in the MS. We propose a data enlarging strategy alongside a novel ensembling strategy that proves to be beneficial for paranasal anomaly classification in the MS.
[ { "version": "v1", "created": "Fri, 31 Mar 2023 09:23:27 GMT" } ]
2023-04-03T00:00:00
[ [ "Bhattacharya", "Debayan", "" ], [ "Behrendt", "Finn", "" ], [ "Becker", "Benjamin Tobias", "" ], [ "Beyersdorff", "Dirk", "" ], [ "Petersen", "Elina", "" ], [ "Petersen", "Marvin", "" ], [ "Cheng", "Bastian", "" ], [ "Eggert", "Dennis", "" ], [ "Betz", "Christian", "" ], [ "Hoffmann", "Anna Sophie", "" ], [ "Schlaefer", "Alexander", "" ] ]
TITLE: Multiple Instance Ensembling For Paranasal Anomaly Classification In The Maxillary Sinus ABSTRACT: Paranasal anomalies are commonly discovered during routine radiological screenings and can present with a wide range of morphological features. This diversity can make it difficult for convolutional neural networks (CNNs) to accurately classify these anomalies, especially when working with limited datasets. Additionally, current approaches to paranasal anomaly classification are constrained to identifying a single anomaly at a time. These challenges necessitate the need for further research and development in this area. In this study, we investigate the feasibility of using a 3D convolutional neural network (CNN) to classify healthy maxillary sinuses (MS) and MS with polyps or cysts. The task of accurately identifying the relevant MS volume within larger head and neck Magnetic Resonance Imaging (MRI) scans can be difficult, but we develop a straightforward strategy to tackle this challenge. Our end-to-end solution includes the use of a novel sampling technique that not only effectively localizes the relevant MS volume, but also increases the size of the training dataset and improves classification results. Additionally, we employ a multiple instance ensemble prediction method to further boost classification performance. Finally, we identify the optimal size of MS volumes to achieve the highest possible classification performance on our dataset. With our multiple instance ensemble prediction strategy and sampling strategy, our 3D CNNs achieve an F1 of 0.85 whereas without it, they achieve an F1 of 0.70. We demonstrate the feasibility of classifying anomalies in the MS. We propose a data enlarging strategy alongside a novel ensembling strategy that proves to be beneficial for paranasal anomaly classification in the MS.
no_new_dataset
0.710396
2303.17921
Haotian Hu
Hu Haotian, Wang Fanyi, Su Jingwen, Gao Shiyu, Zhang Zhiwang
IC-FPS: Instance-Centroid Faster Point Sampling Module for 3D Point-base Object Detection
null
null
null
null
cs.CV
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
3D object detection is one of the most important tasks in autonomous driving and robotics. Our research focuses on tackling low efficiency issue of point-based methods on large-scale point clouds. Existing point-based methods adopt farthest point sampling (FPS) strategy for downsampling, which is computationally expensive in terms of inference time and memory consumption when the number of point cloud increases. In order to improve efficiency, we propose a novel Instance-Centroid Faster Point Sampling Module (IC-FPS) , which effectively replaces the first Set Abstraction (SA) layer that is extremely tedious. IC-FPS module is comprised of two methods, local feature diffusion based background point filter (LFDBF) and Centroid-Instance Sampling Strategy (CISS). LFDBF is constructed to exclude most invalid background points, while CISS substitutes FPS strategy by fast sampling centroids and instance points. IC-FPS module can be inserted to almost every point-based models. Extensive experiments on multiple public benchmarks have demonstrated the superiority of IC-FPS. On Waymo dataset, the proposed module significantly improves performance of baseline model and accelerates inference speed by 3.8 times. For the first time, real-time detection of point-based models in large-scale point cloud scenario is realized.
[ { "version": "v1", "created": "Fri, 31 Mar 2023 09:31:29 GMT" } ]
2023-04-03T00:00:00
[ [ "Haotian", "Hu", "" ], [ "Fanyi", "Wang", "" ], [ "Jingwen", "Su", "" ], [ "Shiyu", "Gao", "" ], [ "Zhiwang", "Zhang", "" ] ]
TITLE: IC-FPS: Instance-Centroid Faster Point Sampling Module for 3D Point-base Object Detection ABSTRACT: 3D object detection is one of the most important tasks in autonomous driving and robotics. Our research focuses on tackling low efficiency issue of point-based methods on large-scale point clouds. Existing point-based methods adopt farthest point sampling (FPS) strategy for downsampling, which is computationally expensive in terms of inference time and memory consumption when the number of point cloud increases. In order to improve efficiency, we propose a novel Instance-Centroid Faster Point Sampling Module (IC-FPS) , which effectively replaces the first Set Abstraction (SA) layer that is extremely tedious. IC-FPS module is comprised of two methods, local feature diffusion based background point filter (LFDBF) and Centroid-Instance Sampling Strategy (CISS). LFDBF is constructed to exclude most invalid background points, while CISS substitutes FPS strategy by fast sampling centroids and instance points. IC-FPS module can be inserted to almost every point-based models. Extensive experiments on multiple public benchmarks have demonstrated the superiority of IC-FPS. On Waymo dataset, the proposed module significantly improves performance of baseline model and accelerates inference speed by 3.8 times. For the first time, real-time detection of point-based models in large-scale point cloud scenario is realized.
no_new_dataset
0.709831
2303.17930
Shiyao Wu
Shiyao Wu
JobHam-place with smart recommend job options and candidate filtering options
null
null
null
null
cs.CL
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
Due to the increasing number of graduates, many applicants experience the situation about finding a job, and employers experience difficulty filtering job applicants, which might negatively impact their effectiveness. However, most job-hunting websites lack job recommendation and CV filtering or ranking functionality, which are not integrated into the system. Thus, a smart job hunter combined with the above functionality will be conducted in this project, which contains job recommendations, CV ranking and even a job dashboard for skills and job applicant functionality. Job recommendation and CV ranking starts from the automatic keyword extraction and end with the Job/CV ranking algorithm. Automatic keyword extraction is implemented by Job2Skill and the CV2Skill model based on Bert. Job2Skill consists of two components, text encoder and Gru-based layers, while CV2Skill is mainly based on Bert and fine-tunes the pre-trained model by the Resume- Entity dataset. Besides, to match skills from CV and job description and rank lists of jobs and candidates, job/CV ranking algorithms have been provided to compute the occurrence ratio of skill words based on TFIDF score and match ratio of the total skill numbers. Besides, some advanced features have been integrated into the website to improve user experiences, such as the calendar and sweetalert2 plugin. And some basic features to go through job application processes, such as job application tracking and interview arrangement.
[ { "version": "v1", "created": "Fri, 31 Mar 2023 09:54:47 GMT" } ]
2023-04-03T00:00:00
[ [ "Wu", "Shiyao", "" ] ]
TITLE: JobHam-place with smart recommend job options and candidate filtering options ABSTRACT: Due to the increasing number of graduates, many applicants experience the situation about finding a job, and employers experience difficulty filtering job applicants, which might negatively impact their effectiveness. However, most job-hunting websites lack job recommendation and CV filtering or ranking functionality, which are not integrated into the system. Thus, a smart job hunter combined with the above functionality will be conducted in this project, which contains job recommendations, CV ranking and even a job dashboard for skills and job applicant functionality. Job recommendation and CV ranking starts from the automatic keyword extraction and end with the Job/CV ranking algorithm. Automatic keyword extraction is implemented by Job2Skill and the CV2Skill model based on Bert. Job2Skill consists of two components, text encoder and Gru-based layers, while CV2Skill is mainly based on Bert and fine-tunes the pre-trained model by the Resume- Entity dataset. Besides, to match skills from CV and job description and rank lists of jobs and candidates, job/CV ranking algorithms have been provided to compute the occurrence ratio of skill words based on TFIDF score and match ratio of the total skill numbers. Besides, some advanced features have been integrated into the website to improve user experiences, such as the calendar and sweetalert2 plugin. And some basic features to go through job application processes, such as job application tracking and interview arrangement.
no_new_dataset
0.710465
2303.17932
Frederic Blum
Frederic Blum and Johann-Mattis List
Trimming Phonetic Alignments Improves the Inference of Sound Correspondence Patterns from Multilingual Wordlists
The paper was accepted at the SIGTYP workshop 2023 co-located with EACL
null
null
null
cs.CL
http://creativecommons.org/licenses/by/4.0/
Sound correspondence patterns form the basis of cognate detection and phonological reconstruction in historical language comparison. Methods for the automatic inference of correspondence patterns from phonetically aligned cognate sets have been proposed, but their application to multilingual wordlists requires extremely well annotated datasets. Since annotation is tedious and time consuming, it would be desirable to find ways to improve aligned cognate data automatically. Taking inspiration from trimming techniques in evolutionary biology, which improve alignments by excluding problematic sites, we propose a workflow that trims phonetic alignments in comparative linguistics prior to the inference of correspondence patterns. Testing these techniques on a large standardized collection of ten datasets with expert annotations from different language families, we find that the best trimming technique substantially improves the overall consistency of the alignments. The results show a clear increase in the proportion of frequent correspondence patterns and words exhibiting regular cognate relations.
[ { "version": "v1", "created": "Fri, 31 Mar 2023 09:55:48 GMT" } ]
2023-04-03T00:00:00
[ [ "Blum", "Frederic", "" ], [ "List", "Johann-Mattis", "" ] ]
TITLE: Trimming Phonetic Alignments Improves the Inference of Sound Correspondence Patterns from Multilingual Wordlists ABSTRACT: Sound correspondence patterns form the basis of cognate detection and phonological reconstruction in historical language comparison. Methods for the automatic inference of correspondence patterns from phonetically aligned cognate sets have been proposed, but their application to multilingual wordlists requires extremely well annotated datasets. Since annotation is tedious and time consuming, it would be desirable to find ways to improve aligned cognate data automatically. Taking inspiration from trimming techniques in evolutionary biology, which improve alignments by excluding problematic sites, we propose a workflow that trims phonetic alignments in comparative linguistics prior to the inference of correspondence patterns. Testing these techniques on a large standardized collection of ten datasets with expert annotations from different language families, we find that the best trimming technique substantially improves the overall consistency of the alignments. The results show a clear increase in the proportion of frequent correspondence patterns and words exhibiting regular cognate relations.
no_new_dataset
0.711994
2303.17937
Xun Xu
Yijin Chen, Xun Xu, Yongyi Su, Kui Jia
STFAR: Improving Object Detection Robustness at Test-Time by Self-Training with Feature Alignment Regularization
null
null
null
null
cs.CV
http://creativecommons.org/licenses/by/4.0/
Domain adaptation helps generalizing object detection models to target domain data with distribution shift. It is often achieved by adapting with access to the whole target domain data. In a more realistic scenario, target distribution is often unpredictable until inference stage. This motivates us to explore adapting an object detection model at test-time, a.k.a. test-time adaptation (TTA). In this work, we approach test-time adaptive object detection (TTAOD) from two perspective. First, we adopt a self-training paradigm to generate pseudo labeled objects with an exponential moving average model. The pseudo labels are further used to supervise adapting source domain model. As self-training is prone to incorrect pseudo labels, we further incorporate aligning feature distributions at two output levels as regularizations to self-training. To validate the performance on TTAOD, we create benchmarks based on three standard object detection datasets and adapt generic TTA methods to object detection task. Extensive evaluations suggest our proposed method sets the state-of-the-art on test-time adaptive object detection task.
[ { "version": "v1", "created": "Fri, 31 Mar 2023 10:04:44 GMT" } ]
2023-04-03T00:00:00
[ [ "Chen", "Yijin", "" ], [ "Xu", "Xun", "" ], [ "Su", "Yongyi", "" ], [ "Jia", "Kui", "" ] ]
TITLE: STFAR: Improving Object Detection Robustness at Test-Time by Self-Training with Feature Alignment Regularization ABSTRACT: Domain adaptation helps generalizing object detection models to target domain data with distribution shift. It is often achieved by adapting with access to the whole target domain data. In a more realistic scenario, target distribution is often unpredictable until inference stage. This motivates us to explore adapting an object detection model at test-time, a.k.a. test-time adaptation (TTA). In this work, we approach test-time adaptive object detection (TTAOD) from two perspective. First, we adopt a self-training paradigm to generate pseudo labeled objects with an exponential moving average model. The pseudo labels are further used to supervise adapting source domain model. As self-training is prone to incorrect pseudo labels, we further incorporate aligning feature distributions at two output levels as regularizations to self-training. To validate the performance on TTAOD, we create benchmarks based on three standard object detection datasets and adapt generic TTA methods to object detection task. Extensive evaluations suggest our proposed method sets the state-of-the-art on test-time adaptive object detection task.
no_new_dataset
0.709239
2303.17942
Bruno Casella
Bruno Casella, Roberto Esposito, Carlo Cavazzoni, Marco Aldinucci
Benchmarking FedAvg and FedCurv for Image Classification Tasks
12 pages, Proceedings of ITADATA22, The 1st Italian Conference on Big Data and Data Science; Published on CEUR Workshop Proceedings (CEUR-WS.org, ISSN 1613-0073), Vol. 3340, pp. 99-110, 2022
CEUR Workshop Proceedings Vol. 3340, pp. 99-110, (2022)
null
null
cs.LG cs.CV cs.DC
http://creativecommons.org/licenses/by/4.0/
Classic Machine Learning techniques require training on data available in a single data lake. However, aggregating data from different owners is not always convenient for different reasons, including security, privacy and secrecy. Data carry a value that might vanish when shared with others; the ability to avoid sharing the data enables industrial applications where security and privacy are of paramount importance, making it possible to train global models by implementing only local policies which can be run independently and even on air-gapped data centres. Federated Learning (FL) is a distributed machine learning approach which has emerged as an effective way to address privacy concerns by only sharing local AI models while keeping the data decentralized. Two critical challenges of Federated Learning are managing the heterogeneous systems in the same federated network and dealing with real data, which are often not independently and identically distributed (non-IID) among the clients. In this paper, we focus on the second problem, i.e., the problem of statistical heterogeneity of the data in the same federated network. In this setting, local models might be strayed far from the local optimum of the complete dataset, thus possibly hindering the convergence of the federated model. Several Federated Learning algorithms, such as FedAvg, FedProx and Federated Curvature (FedCurv), aiming at tackling the non-IID setting, have already been proposed. This work provides an empirical assessment of the behaviour of FedAvg and FedCurv in common non-IID scenarios. Results show that the number of epochs per round is an important hyper-parameter that, when tuned appropriately, can lead to significant performance gains while reducing the communication cost. As a side product of this work, we release the non-IID version of the datasets we used so to facilitate further comparisons from the FL community.
[ { "version": "v1", "created": "Fri, 31 Mar 2023 10:13:01 GMT" } ]
2023-04-03T00:00:00
[ [ "Casella", "Bruno", "" ], [ "Esposito", "Roberto", "" ], [ "Cavazzoni", "Carlo", "" ], [ "Aldinucci", "Marco", "" ] ]
TITLE: Benchmarking FedAvg and FedCurv for Image Classification Tasks ABSTRACT: Classic Machine Learning techniques require training on data available in a single data lake. However, aggregating data from different owners is not always convenient for different reasons, including security, privacy and secrecy. Data carry a value that might vanish when shared with others; the ability to avoid sharing the data enables industrial applications where security and privacy are of paramount importance, making it possible to train global models by implementing only local policies which can be run independently and even on air-gapped data centres. Federated Learning (FL) is a distributed machine learning approach which has emerged as an effective way to address privacy concerns by only sharing local AI models while keeping the data decentralized. Two critical challenges of Federated Learning are managing the heterogeneous systems in the same federated network and dealing with real data, which are often not independently and identically distributed (non-IID) among the clients. In this paper, we focus on the second problem, i.e., the problem of statistical heterogeneity of the data in the same federated network. In this setting, local models might be strayed far from the local optimum of the complete dataset, thus possibly hindering the convergence of the federated model. Several Federated Learning algorithms, such as FedAvg, FedProx and Federated Curvature (FedCurv), aiming at tackling the non-IID setting, have already been proposed. This work provides an empirical assessment of the behaviour of FedAvg and FedCurv in common non-IID scenarios. Results show that the number of epochs per round is an important hyper-parameter that, when tuned appropriately, can lead to significant performance gains while reducing the communication cost. As a side product of this work, we release the non-IID version of the datasets we used so to facilitate further comparisons from the FL community.
no_new_dataset
0.707771
2303.17948
Ming Yan
Ming Yan, Xin Wang, Yudi Dai, Siqi Shen, Chenglu Wen, Lan Xu, Yuexin Ma, Cheng Wang
CIMI4D: A Large Multimodal Climbing Motion Dataset under Human-scene Interactions
CVPR 2023
null
null
null
cs.CV
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
Motion capture is a long-standing research problem. Although it has been studied for decades, the majority of research focus on ground-based movements such as walking, sitting, dancing, etc. Off-grounded actions such as climbing are largely overlooked. As an important type of action in sports and firefighting field, the climbing movements is challenging to capture because of its complex back poses, intricate human-scene interactions, and difficult global localization. The research community does not have an in-depth understanding of the climbing action due to the lack of specific datasets. To address this limitation, we collect CIMI4D, a large rock \textbf{C}l\textbf{I}mbing \textbf{M}ot\textbf{I}on dataset from 12 persons climbing 13 different climbing walls. The dataset consists of around 180,000 frames of pose inertial measurements, LiDAR point clouds, RGB videos, high-precision static point cloud scenes, and reconstructed scene meshes. Moreover, we frame-wise annotate touch rock holds to facilitate a detailed exploration of human-scene interaction. The core of this dataset is a blending optimization process, which corrects for the pose as it drifts and is affected by the magnetic conditions. To evaluate the merit of CIMI4D, we perform four tasks which include human pose estimations (with/without scene constraints), pose prediction, and pose generation. The experimental results demonstrate that CIMI4D presents great challenges to existing methods and enables extensive research opportunities. We share the dataset with the research community in http://www.lidarhumanmotion.net/cimi4d/.
[ { "version": "v1", "created": "Fri, 31 Mar 2023 10:26:47 GMT" } ]
2023-04-03T00:00:00
[ [ "Yan", "Ming", "" ], [ "Wang", "Xin", "" ], [ "Dai", "Yudi", "" ], [ "Shen", "Siqi", "" ], [ "Wen", "Chenglu", "" ], [ "Xu", "Lan", "" ], [ "Ma", "Yuexin", "" ], [ "Wang", "Cheng", "" ] ]
TITLE: CIMI4D: A Large Multimodal Climbing Motion Dataset under Human-scene Interactions ABSTRACT: Motion capture is a long-standing research problem. Although it has been studied for decades, the majority of research focus on ground-based movements such as walking, sitting, dancing, etc. Off-grounded actions such as climbing are largely overlooked. As an important type of action in sports and firefighting field, the climbing movements is challenging to capture because of its complex back poses, intricate human-scene interactions, and difficult global localization. The research community does not have an in-depth understanding of the climbing action due to the lack of specific datasets. To address this limitation, we collect CIMI4D, a large rock \textbf{C}l\textbf{I}mbing \textbf{M}ot\textbf{I}on dataset from 12 persons climbing 13 different climbing walls. The dataset consists of around 180,000 frames of pose inertial measurements, LiDAR point clouds, RGB videos, high-precision static point cloud scenes, and reconstructed scene meshes. Moreover, we frame-wise annotate touch rock holds to facilitate a detailed exploration of human-scene interaction. The core of this dataset is a blending optimization process, which corrects for the pose as it drifts and is affected by the magnetic conditions. To evaluate the merit of CIMI4D, we perform four tasks which include human pose estimations (with/without scene constraints), pose prediction, and pose generation. The experimental results demonstrate that CIMI4D presents great challenges to existing methods and enables extensive research opportunities. We share the dataset with the research community in http://www.lidarhumanmotion.net/cimi4d/.
new_dataset
0.719738
2303.17949
Anbai Jiang
Anbai Jiang, Wei-Qiang Zhang, Yufeng Deng, Pingyi Fan and Jia Liu
Unsupervised Anomaly Detection and Localization of Machine Audio: A GAN-based Approach
Accepted by ICASSP 2023
null
null
null
cs.SD cs.LG eess.AS
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
Automatic detection of machine anomaly remains challenging for machine learning. We believe the capability of generative adversarial network (GAN) suits the need of machine audio anomaly detection, yet rarely has this been investigated by previous work. In this paper, we propose AEGAN-AD, a totally unsupervised approach in which the generator (also an autoencoder) is trained to reconstruct input spectrograms. It is pointed out that the denoising nature of reconstruction deprecates its capacity. Thus, the discriminator is redesigned to aid the generator during both training stage and detection stage. The performance of AEGAN-AD on the dataset of DCASE 2022 Challenge TASK 2 demonstrates the state-of-the-art result on five machine types. A novel anomaly localization method is also investigated. Source code available at: www.github.com/jianganbai/AEGAN-AD
[ { "version": "v1", "created": "Fri, 31 Mar 2023 10:27:36 GMT" } ]
2023-04-03T00:00:00
[ [ "Jiang", "Anbai", "" ], [ "Zhang", "Wei-Qiang", "" ], [ "Deng", "Yufeng", "" ], [ "Fan", "Pingyi", "" ], [ "Liu", "Jia", "" ] ]
TITLE: Unsupervised Anomaly Detection and Localization of Machine Audio: A GAN-based Approach ABSTRACT: Automatic detection of machine anomaly remains challenging for machine learning. We believe the capability of generative adversarial network (GAN) suits the need of machine audio anomaly detection, yet rarely has this been investigated by previous work. In this paper, we propose AEGAN-AD, a totally unsupervised approach in which the generator (also an autoencoder) is trained to reconstruct input spectrograms. It is pointed out that the denoising nature of reconstruction deprecates its capacity. Thus, the discriminator is redesigned to aid the generator during both training stage and detection stage. The performance of AEGAN-AD on the dataset of DCASE 2022 Challenge TASK 2 demonstrates the state-of-the-art result on five machine types. A novel anomaly localization method is also investigated. Source code available at: www.github.com/jianganbai/AEGAN-AD
no_new_dataset
0.708855
2303.17966
Kang Li
Zhi Yang, Kang Li, Haitao Gan, Zhongwei Huang, Ming Shi
HD-GCN:A Hybrid Diffusion Graph Convolutional Network
16 pages, 4 figures
null
null
null
cs.LG
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
The information diffusion performance of GCN and its variant models is limited by the adjacency matrix, which can lower their performance. Therefore, we introduce a new framework for graph convolutional networks called Hybrid Diffusion-based Graph Convolutional Network (HD-GCN) to address the limitations of information diffusion caused by the adjacency matrix. In the HD-GCN framework, we initially utilize diffusion maps to facilitate the diffusion of information among nodes that are adjacent to each other in the feature space. This allows for the diffusion of information between similar points that may not have an adjacent relationship. Next, we utilize graph convolution to further propagate information among adjacent nodes after the diffusion maps, thereby enabling the spread of information among similar nodes that are adjacent in the graph. Finally, we employ the diffusion distances obtained through the use of diffusion maps to regularize and constrain the predicted labels of training nodes. This regularization method is then applied to the HD-GCN training, resulting in a smoother classification surface. The model proposed in this paper effectively overcomes the limitations of information diffusion imposed only by the adjacency matrix. HD-GCN utilizes hybrid diffusion by combining information diffusion between neighborhood nodes in the feature space and adjacent nodes in the adjacency matrix. This method allows for more comprehensive information propagation among nodes, resulting in improved model performance. We evaluated the performance of DM-GCN on three well-known citation network datasets and the results showed that the proposed framework is more effective than several graph-based semi-supervised learning methods.
[ { "version": "v1", "created": "Fri, 31 Mar 2023 11:12:25 GMT" } ]
2023-04-03T00:00:00
[ [ "Yang", "Zhi", "" ], [ "Li", "Kang", "" ], [ "Gan", "Haitao", "" ], [ "Huang", "Zhongwei", "" ], [ "Shi", "Ming", "" ] ]
TITLE: HD-GCN:A Hybrid Diffusion Graph Convolutional Network ABSTRACT: The information diffusion performance of GCN and its variant models is limited by the adjacency matrix, which can lower their performance. Therefore, we introduce a new framework for graph convolutional networks called Hybrid Diffusion-based Graph Convolutional Network (HD-GCN) to address the limitations of information diffusion caused by the adjacency matrix. In the HD-GCN framework, we initially utilize diffusion maps to facilitate the diffusion of information among nodes that are adjacent to each other in the feature space. This allows for the diffusion of information between similar points that may not have an adjacent relationship. Next, we utilize graph convolution to further propagate information among adjacent nodes after the diffusion maps, thereby enabling the spread of information among similar nodes that are adjacent in the graph. Finally, we employ the diffusion distances obtained through the use of diffusion maps to regularize and constrain the predicted labels of training nodes. This regularization method is then applied to the HD-GCN training, resulting in a smoother classification surface. The model proposed in this paper effectively overcomes the limitations of information diffusion imposed only by the adjacency matrix. HD-GCN utilizes hybrid diffusion by combining information diffusion between neighborhood nodes in the feature space and adjacent nodes in the adjacency matrix. This method allows for more comprehensive information propagation among nodes, resulting in improved model performance. We evaluated the performance of DM-GCN on three well-known citation network datasets and the results showed that the proposed framework is more effective than several graph-based semi-supervised learning methods.
no_new_dataset
0.712589
2303.17984
Zf Wu
Zifan Wu, Chao Yu, Chen Chen, Jianye Hao, Hankz Hankui Zhuo
Models as Agents: Optimizing Multi-Step Predictions of Interactive Local Models in Model-Based Multi-Agent Reinforcement Learning
null
null
null
null
cs.MA
http://creativecommons.org/licenses/by/4.0/
Research in model-based reinforcement learning has made significant progress in recent years. Compared to single-agent settings, the exponential dimension growth of the joint state-action space in multi-agent systems dramatically increases the complexity of the environment dynamics, which makes it infeasible to learn an accurate global model and thus necessitates the use of agent-wise local models. However, during multi-step model rollouts, the prediction of one local model can affect the predictions of other local models in the next step. As a result, local prediction errors can be propagated to other localities and eventually give rise to considerably large global errors. Furthermore, since the models are generally used to predict for multiple steps, simply minimizing one-step prediction errors regardless of their long-term effect on other models may further aggravate the propagation of local errors. To this end, we propose Models as AGents (MAG), a multi-agent model optimization framework that reversely treats the local models as multi-step decision making agents and the current policies as the dynamics during the model rollout process. In this way, the local models are able to consider the multi-step mutual affect between each other before making predictions. Theoretically, we show that the objective of MAG is approximately equivalent to maximizing a lower bound of the true environment return. Experiments on the challenging StarCraft II benchmark demonstrate the effectiveness of MAG.
[ { "version": "v1", "created": "Fri, 31 Mar 2023 11:42:04 GMT" } ]
2023-04-03T00:00:00
[ [ "Wu", "Zifan", "" ], [ "Yu", "Chao", "" ], [ "Chen", "Chen", "" ], [ "Hao", "Jianye", "" ], [ "Zhuo", "Hankz Hankui", "" ] ]
TITLE: Models as Agents: Optimizing Multi-Step Predictions of Interactive Local Models in Model-Based Multi-Agent Reinforcement Learning ABSTRACT: Research in model-based reinforcement learning has made significant progress in recent years. Compared to single-agent settings, the exponential dimension growth of the joint state-action space in multi-agent systems dramatically increases the complexity of the environment dynamics, which makes it infeasible to learn an accurate global model and thus necessitates the use of agent-wise local models. However, during multi-step model rollouts, the prediction of one local model can affect the predictions of other local models in the next step. As a result, local prediction errors can be propagated to other localities and eventually give rise to considerably large global errors. Furthermore, since the models are generally used to predict for multiple steps, simply minimizing one-step prediction errors regardless of their long-term effect on other models may further aggravate the propagation of local errors. To this end, we propose Models as AGents (MAG), a multi-agent model optimization framework that reversely treats the local models as multi-step decision making agents and the current policies as the dynamics during the model rollout process. In this way, the local models are able to consider the multi-step mutual affect between each other before making predictions. Theoretically, we show that the objective of MAG is approximately equivalent to maximizing a lower bound of the true environment return. Experiments on the challenging StarCraft II benchmark demonstrate the effectiveness of MAG.
no_new_dataset
0.708975
2303.17989
Panagiotis Agrafiotis
Panagiotis Agrafiotis, Anastastios Doulamis, Andreas Georgopoulos
Unsupervised crack detection on complex stone masonry surfaces
Submitted to the Journal of Cultural Heritage, Elsevier, under review as of 31st of March 2023
null
null
null
cs.CV
http://creativecommons.org/licenses/by-nc-sa/4.0/
Computer vision for detecting building pathologies has interested researchers for quite some time. Vision-based crack detection is a non-destructive assessment technique, which can be useful especially for Cultural Heritage (CH) where strict regulations apply and, even simple, interventions are not permitted. Recently, shallow and deep machine learning architectures applied on various types of imagery are gaining ground. In this article a crack detection methodology for stone masonry walls is presented. In the proposed approach, crack detection is approached as an unsupervised anomaly detection problem on RGB (Red Green Blue) image patches. Towards this direction, some of the most popular state of the art CNN (Convolutional Neural Network) architectures are deployed and modified to binary classify the images or image patches by predicting a specific class for the tested imagery; 'Crack' or 'No crack', and detect and localize those cracks on the RGB imagery with high accuracy. Testing of the model was performed on various test sites and random images retrieved from the internet and collected by the authors and results suggested the high performance of specific networks compared to the rest, considering also the small numbers of epochs required for training. Those results met the accuracy delivered by more complex and computationally heavy approaches, requiring a large amount of data for training. Source code is available on GitHub https://github.com/pagraf/Crack-detection while datasets are available on Zenodo https://doi.org/10.5281/zenodo.6516913 .
[ { "version": "v1", "created": "Fri, 31 Mar 2023 12:07:23 GMT" } ]
2023-04-03T00:00:00
[ [ "Agrafiotis", "Panagiotis", "" ], [ "Doulamis", "Anastastios", "" ], [ "Georgopoulos", "Andreas", "" ] ]
TITLE: Unsupervised crack detection on complex stone masonry surfaces ABSTRACT: Computer vision for detecting building pathologies has interested researchers for quite some time. Vision-based crack detection is a non-destructive assessment technique, which can be useful especially for Cultural Heritage (CH) where strict regulations apply and, even simple, interventions are not permitted. Recently, shallow and deep machine learning architectures applied on various types of imagery are gaining ground. In this article a crack detection methodology for stone masonry walls is presented. In the proposed approach, crack detection is approached as an unsupervised anomaly detection problem on RGB (Red Green Blue) image patches. Towards this direction, some of the most popular state of the art CNN (Convolutional Neural Network) architectures are deployed and modified to binary classify the images or image patches by predicting a specific class for the tested imagery; 'Crack' or 'No crack', and detect and localize those cracks on the RGB imagery with high accuracy. Testing of the model was performed on various test sites and random images retrieved from the internet and collected by the authors and results suggested the high performance of specific networks compared to the rest, considering also the small numbers of epochs required for training. Those results met the accuracy delivered by more complex and computationally heavy approaches, requiring a large amount of data for training. Source code is available on GitHub https://github.com/pagraf/Crack-detection while datasets are available on Zenodo https://doi.org/10.5281/zenodo.6516913 .
no_new_dataset
0.711794
2303.18001
A A
Zhaoxu Li, Yingqian Wang, Chao Xiao, Qiang Ling, Zaiping Lin, and Wei An
You Only Train Once: Learning a General Anomaly Enhancement Network with Random Masks for Hyperspectral Anomaly Detection
null
TGRS 2023
10.1109/TGRS.2023.3258067
null
eess.IV cs.CV
http://creativecommons.org/licenses/by/4.0/
In this paper, we introduce a new approach to address the challenge of generalization in hyperspectral anomaly detection (AD). Our method eliminates the need for adjusting parameters or retraining on new test scenes as required by most existing methods. Employing an image-level training paradigm, we achieve a general anomaly enhancement network for hyperspectral AD that only needs to be trained once. Trained on a set of anomaly-free hyperspectral images with random masks, our network can learn the spatial context characteristics between anomalies and background in an unsupervised way. Additionally, a plug-and-play model selection module is proposed to search for a spatial-spectral transform domain that is more suitable for AD task than the original data. To establish a unified benchmark to comprehensively evaluate our method and existing methods, we develop a large-scale hyperspectral AD dataset (HAD100) that includes 100 real test scenes with diverse anomaly targets. In comparison experiments, we combine our network with a parameter-free detector and achieve the optimal balance between detection accuracy and inference speed among state-of-the-art AD methods. Experimental results also show that our method still achieves competitive performance when the training and test set are captured by different sensor devices. Our code is available at https://github.com/ZhaoxuLi123/AETNet.
[ { "version": "v1", "created": "Fri, 31 Mar 2023 12:23:56 GMT" } ]
2023-04-03T00:00:00
[ [ "Li", "Zhaoxu", "" ], [ "Wang", "Yingqian", "" ], [ "Xiao", "Chao", "" ], [ "Ling", "Qiang", "" ], [ "Lin", "Zaiping", "" ], [ "An", "Wei", "" ] ]
TITLE: You Only Train Once: Learning a General Anomaly Enhancement Network with Random Masks for Hyperspectral Anomaly Detection ABSTRACT: In this paper, we introduce a new approach to address the challenge of generalization in hyperspectral anomaly detection (AD). Our method eliminates the need for adjusting parameters or retraining on new test scenes as required by most existing methods. Employing an image-level training paradigm, we achieve a general anomaly enhancement network for hyperspectral AD that only needs to be trained once. Trained on a set of anomaly-free hyperspectral images with random masks, our network can learn the spatial context characteristics between anomalies and background in an unsupervised way. Additionally, a plug-and-play model selection module is proposed to search for a spatial-spectral transform domain that is more suitable for AD task than the original data. To establish a unified benchmark to comprehensively evaluate our method and existing methods, we develop a large-scale hyperspectral AD dataset (HAD100) that includes 100 real test scenes with diverse anomaly targets. In comparison experiments, we combine our network with a parameter-free detector and achieve the optimal balance between detection accuracy and inference speed among state-of-the-art AD methods. Experimental results also show that our method still achieves competitive performance when the training and test set are captured by different sensor devices. Our code is available at https://github.com/ZhaoxuLi123/AETNet.
new_dataset
0.712782
2303.18006
Abhishek Gupta
Emily Dardaman (1) and Abhishek Gupta (1, 2, and 3) ((1) BCG Henderson Institute, (2) Montreal AI Ethics Institute, (3) Boston Consulting Group)
Asking Better Questions -- The Art and Science of Forecasting: A mechanism for truer answers to high-stakes questions
7 pages
null
null
null
cs.CY cs.AI
http://creativecommons.org/licenses/by/4.0/
Without the ability to estimate and benchmark AI capability advancements, organizations are left to respond to each change reactively, impeding their ability to build viable mid and long-term strategies. This paper explores the recent growth of forecasting, a political science tool that uses explicit assumptions and quantitative estimation that leads to improved prediction accuracy. Done at the collective level, forecasting can identify and verify talent, enable leaders to build better models of AI advancements and improve inputs into design policy. Successful approaches to forecasting and case studies are examined, revealing a subclass of "superforecasters" who outperform 98% of the population and whose insights will be most reliable. Finally, techniques behind successful forecasting are outlined, including Phillip Tetlock's "Ten Commandments." To adapt to a quickly changing technology landscape, designers and policymakers should consider forecasting as a first line of defense.
[ { "version": "v1", "created": "Fri, 31 Mar 2023 12:27:26 GMT" } ]
2023-04-03T00:00:00
[ [ "Dardaman", "Emily", "", "1, 2, and 3" ], [ "Gupta", "Abhishek", "", "1, 2, and 3" ] ]
TITLE: Asking Better Questions -- The Art and Science of Forecasting: A mechanism for truer answers to high-stakes questions ABSTRACT: Without the ability to estimate and benchmark AI capability advancements, organizations are left to respond to each change reactively, impeding their ability to build viable mid and long-term strategies. This paper explores the recent growth of forecasting, a political science tool that uses explicit assumptions and quantitative estimation that leads to improved prediction accuracy. Done at the collective level, forecasting can identify and verify talent, enable leaders to build better models of AI advancements and improve inputs into design policy. Successful approaches to forecasting and case studies are examined, revealing a subclass of "superforecasters" who outperform 98% of the population and whose insights will be most reliable. Finally, techniques behind successful forecasting are outlined, including Phillip Tetlock's "Ten Commandments." To adapt to a quickly changing technology landscape, designers and policymakers should consider forecasting as a first line of defense.
no_new_dataset
0.7076
2303.18022
Angel Victor Juanco Muller
Angel Victor Juanco Muller, Joao F.C. Mota, Keith A. Goatman, Corne Hoogendoorn
The Topology-Overlap Trade-Off in Retinal Arteriole-Venule Segmentation
To be published in proceedings of SPIE Medical Imaging 2023 Image Processing
null
null
null
cs.CV stat.ML
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
Retinal fundus images can be an invaluable diagnosis tool for screening epidemic diseases like hypertension or diabetes. And they become especially useful when the arterioles and venules they depict are clearly identified and annotated. However, manual annotation of these vessels is extremely time demanding and taxing, which calls for automatic segmentation. Although convolutional neural networks can achieve high overlap between predictions and expert annotations, they often fail to produce topologically correct predictions of tubular structures. This situation is exacerbated by the bifurcation versus crossing ambiguity which causes classification mistakes. This paper shows that including a topology preserving term in the loss function improves the continuity of the segmented vessels, although at the expense of artery-vein misclassification and overall lower overlap metrics. However, we show that by including an orientation score guided convolutional module, based on the anisotropic single sided cake wavelet, we reduce such misclassification and further increase the topology correctness of the results. We evaluate our model on public datasets with conveniently chosen metrics to assess both overlap and topology correctness, showing that our model is able to produce results on par with state-of-the-art from the point of view of overlap, while increasing topological accuracy.
[ { "version": "v1", "created": "Fri, 31 Mar 2023 13:01:05 GMT" } ]
2023-04-03T00:00:00
[ [ "Muller", "Angel Victor Juanco", "" ], [ "Mota", "Joao F. C.", "" ], [ "Goatman", "Keith A.", "" ], [ "Hoogendoorn", "Corne", "" ] ]
TITLE: The Topology-Overlap Trade-Off in Retinal Arteriole-Venule Segmentation ABSTRACT: Retinal fundus images can be an invaluable diagnosis tool for screening epidemic diseases like hypertension or diabetes. And they become especially useful when the arterioles and venules they depict are clearly identified and annotated. However, manual annotation of these vessels is extremely time demanding and taxing, which calls for automatic segmentation. Although convolutional neural networks can achieve high overlap between predictions and expert annotations, they often fail to produce topologically correct predictions of tubular structures. This situation is exacerbated by the bifurcation versus crossing ambiguity which causes classification mistakes. This paper shows that including a topology preserving term in the loss function improves the continuity of the segmented vessels, although at the expense of artery-vein misclassification and overall lower overlap metrics. However, we show that by including an orientation score guided convolutional module, based on the anisotropic single sided cake wavelet, we reduce such misclassification and further increase the topology correctness of the results. We evaluate our model on public datasets with conveniently chosen metrics to assess both overlap and topology correctness, showing that our model is able to produce results on par with state-of-the-art from the point of view of overlap, while increasing topological accuracy.
no_new_dataset
0.711838
2303.18037
Jingzhan Ge
Jingzhan Ge
Traffic Sign Recognition Dataset and Data Augmentation
14pages, 11 figures
null
null
null
cs.CV cs.LG
http://creativecommons.org/licenses/by/4.0/
Although there are many datasets for traffic sign classification, there are few datasets collected for traffic sign recognition and few of them obtain enough instances especially for training a model with the deep learning method. The deep learning method is almost the only way to train a model for real-world usage that covers various highly similar classes compared with the traditional way such as through color, shape, etc. Also, for some certain sign classes, their sign meanings were destined to can't get enough instances in the dataset. To solve this problem, we purpose a unique data augmentation method for the traffic sign recognition dataset that takes advantage of the standard of the traffic sign. We called it TSR dataset augmentation. We based on the benchmark Tsinghua-Tencent 100K (TT100K) dataset to verify the unique data augmentation method. we performed the method on four main iteration version datasets based on the TT100K dataset and the experimental results showed our method is efficacious. The iteration version datasets based on TT100K, data augmentation method source code and the training results introduced in this paper are publicly available.
[ { "version": "v1", "created": "Fri, 31 Mar 2023 13:14:36 GMT" } ]
2023-04-03T00:00:00
[ [ "Ge", "Jingzhan", "" ] ]
TITLE: Traffic Sign Recognition Dataset and Data Augmentation ABSTRACT: Although there are many datasets for traffic sign classification, there are few datasets collected for traffic sign recognition and few of them obtain enough instances especially for training a model with the deep learning method. The deep learning method is almost the only way to train a model for real-world usage that covers various highly similar classes compared with the traditional way such as through color, shape, etc. Also, for some certain sign classes, their sign meanings were destined to can't get enough instances in the dataset. To solve this problem, we purpose a unique data augmentation method for the traffic sign recognition dataset that takes advantage of the standard of the traffic sign. We called it TSR dataset augmentation. We based on the benchmark Tsinghua-Tencent 100K (TT100K) dataset to verify the unique data augmentation method. we performed the method on four main iteration version datasets based on the TT100K dataset and the experimental results showed our method is efficacious. The iteration version datasets based on TT100K, data augmentation method source code and the training results introduced in this paper are publicly available.
no_new_dataset
0.701342
2303.18084
Chenghao Shi
Chenghao Shi, Xieyuanli Chen, Huimin Lu, Wenbang Deng, Junhao Xiao, Bin Dai
RDMNet: Reliable Dense Matching Based Point Cloud Registration for Autonomous Driving
11 pages, 9 figures
null
null
null
cs.CV cs.RO
http://creativecommons.org/licenses/by/4.0/
Point cloud registration is an important task in robotics and autonomous driving to estimate the ego-motion of the vehicle. Recent advances following the coarse-to-fine manner show promising potential in point cloud registration. However, existing methods rely on good superpoint correspondences, which are hard to be obtained reliably and efficiently, thus resulting in less robust and accurate point cloud registration. In this paper, we propose a novel network, named RDMNet, to find dense point correspondences coarse-to-fine and improve final pose estimation based on such reliable correspondences. Our RDMNet uses a devised 3D-RoFormer mechanism to first extract distinctive superpoints and generates reliable superpoints matches between two point clouds. The proposed 3D-RoFormer fuses 3D position information into the transformer network, efficiently exploiting point clouds' contextual and geometric information to generate robust superpoint correspondences. RDMNet then propagates the sparse superpoints matches to dense point matches using the neighborhood information for accurate point cloud registration. We extensively evaluate our method on multiple datasets from different environments. The experimental results demonstrate that our method outperforms existing state-of-the-art approaches in all tested datasets with a strong generalization ability.
[ { "version": "v1", "created": "Fri, 31 Mar 2023 14:22:32 GMT" } ]
2023-04-03T00:00:00
[ [ "Shi", "Chenghao", "" ], [ "Chen", "Xieyuanli", "" ], [ "Lu", "Huimin", "" ], [ "Deng", "Wenbang", "" ], [ "Xiao", "Junhao", "" ], [ "Dai", "Bin", "" ] ]
TITLE: RDMNet: Reliable Dense Matching Based Point Cloud Registration for Autonomous Driving ABSTRACT: Point cloud registration is an important task in robotics and autonomous driving to estimate the ego-motion of the vehicle. Recent advances following the coarse-to-fine manner show promising potential in point cloud registration. However, existing methods rely on good superpoint correspondences, which are hard to be obtained reliably and efficiently, thus resulting in less robust and accurate point cloud registration. In this paper, we propose a novel network, named RDMNet, to find dense point correspondences coarse-to-fine and improve final pose estimation based on such reliable correspondences. Our RDMNet uses a devised 3D-RoFormer mechanism to first extract distinctive superpoints and generates reliable superpoints matches between two point clouds. The proposed 3D-RoFormer fuses 3D position information into the transformer network, efficiently exploiting point clouds' contextual and geometric information to generate robust superpoint correspondences. RDMNet then propagates the sparse superpoints matches to dense point matches using the neighborhood information for accurate point cloud registration. We extensively evaluate our method on multiple datasets from different environments. The experimental results demonstrate that our method outperforms existing state-of-the-art approaches in all tested datasets with a strong generalization ability.
no_new_dataset
0.711017
2303.18087
Esther Rolf
Esther Rolf
Evaluation Challenges for Geospatial ML
ICLR 2023 Workshop on Machine Learning for Remote Sensing
null
null
null
cs.LG stat.ML
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
As geospatial machine learning models and maps derived from their predictions are increasingly used for downstream analyses in science and policy, it is imperative to evaluate their accuracy and applicability. Geospatial machine learning has key distinctions from other learning paradigms, and as such, the correct way to measure performance of spatial machine learning outputs has been a topic of debate. In this paper, I delineate unique challenges of model evaluation for geospatial machine learning with global or remotely sensed datasets, culminating in concrete takeaways to improve evaluations of geospatial model performance.
[ { "version": "v1", "created": "Fri, 31 Mar 2023 14:24:06 GMT" } ]
2023-04-03T00:00:00
[ [ "Rolf", "Esther", "" ] ]
TITLE: Evaluation Challenges for Geospatial ML ABSTRACT: As geospatial machine learning models and maps derived from their predictions are increasingly used for downstream analyses in science and policy, it is imperative to evaluate their accuracy and applicability. Geospatial machine learning has key distinctions from other learning paradigms, and as such, the correct way to measure performance of spatial machine learning outputs has been a topic of debate. In this paper, I delineate unique challenges of model evaluation for geospatial machine learning with global or remotely sensed datasets, culminating in concrete takeaways to improve evaluations of geospatial model performance.
no_new_dataset
0.713038
2303.18103
B\"orje F. Karlsson
Sanxing Chen, Yongqiang Chen, B\"orje F. Karlsson
Dataset and Baseline System for Multi-lingual Extraction and Normalization of Temporal and Numerical Expressions
Technical Report
null
null
MSR-TR-2023-9
cs.CL cs.AI
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
Temporal and numerical expression understanding is of great importance in many downstream Natural Language Processing (NLP) and Information Retrieval (IR) tasks. However, much previous work covers only a few sub-types and focuses only on entity extraction, which severely limits the usability of identified mentions. In order for such entities to be useful in downstream scenarios, coverage and granularity of sub-types are important; and, even more so, providing resolution into concrete values that can be manipulated. Furthermore, most previous work addresses only a handful of languages. Here we describe a multi-lingual evaluation dataset - NTX - covering diverse temporal and numerical expressions across 14 languages and covering extraction, normalization, and resolution. Along with the dataset we provide a robust rule-based system as a strong baseline for comparisons against other models to be evaluated in this dataset. Data and code are available at \url{https://aka.ms/NTX}.
[ { "version": "v1", "created": "Fri, 31 Mar 2023 14:49:05 GMT" } ]
2023-04-03T00:00:00
[ [ "Chen", "Sanxing", "" ], [ "Chen", "Yongqiang", "" ], [ "Karlsson", "Börje F.", "" ] ]
TITLE: Dataset and Baseline System for Multi-lingual Extraction and Normalization of Temporal and Numerical Expressions ABSTRACT: Temporal and numerical expression understanding is of great importance in many downstream Natural Language Processing (NLP) and Information Retrieval (IR) tasks. However, much previous work covers only a few sub-types and focuses only on entity extraction, which severely limits the usability of identified mentions. In order for such entities to be useful in downstream scenarios, coverage and granularity of sub-types are important; and, even more so, providing resolution into concrete values that can be manipulated. Furthermore, most previous work addresses only a handful of languages. Here we describe a multi-lingual evaluation dataset - NTX - covering diverse temporal and numerical expressions across 14 languages and covering extraction, normalization, and resolution. Along with the dataset we provide a robust rule-based system as a strong baseline for comparisons against other models to be evaluated in this dataset. Data and code are available at \url{https://aka.ms/NTX}.
new_dataset
0.713082
2303.18110
Ramon Sanabria
Ramon Sanabria, Nikolay Bogoychev, Nina Markl, Andrea Carmantini, Ondrej Klejch, Peter Bell
The Edinburgh International Accents of English Corpus: Towards the Democratization of English ASR
Accepted to IEEE ICASSP 2023
null
null
null
cs.CL cs.LG cs.SD eess.AS
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
English is the most widely spoken language in the world, used daily by millions of people as a first or second language in many different contexts. As a result, there are many varieties of English. Although the great many advances in English automatic speech recognition (ASR) over the past decades, results are usually reported based on test datasets which fail to represent the diversity of English as spoken today around the globe. We present the first release of The Edinburgh International Accents of English Corpus (EdAcc). This dataset attempts to better represent the wide diversity of English, encompassing almost 40 hours of dyadic video call conversations between friends. Unlike other datasets, EdAcc includes a wide range of first and second-language varieties of English and a linguistic background profile of each speaker. Results on latest public, and commercial models show that EdAcc highlights shortcomings of current English ASR models. The best performing model, trained on 680 thousand hours of transcribed data, obtains an average of 19.7% word error rate (WER) -- in contrast to the 2.7% WER obtained when evaluated on US English clean read speech. Across all models, we observe a drop in performance on Indian, Jamaican, and Nigerian English speakers. Recordings, linguistic backgrounds, data statement, and evaluation scripts are released on our website (https://groups.inf.ed.ac.uk/edacc/) under CC-BY-SA license.
[ { "version": "v1", "created": "Fri, 31 Mar 2023 14:56:54 GMT" } ]
2023-04-03T00:00:00
[ [ "Sanabria", "Ramon", "" ], [ "Bogoychev", "Nikolay", "" ], [ "Markl", "Nina", "" ], [ "Carmantini", "Andrea", "" ], [ "Klejch", "Ondrej", "" ], [ "Bell", "Peter", "" ] ]
TITLE: The Edinburgh International Accents of English Corpus: Towards the Democratization of English ASR ABSTRACT: English is the most widely spoken language in the world, used daily by millions of people as a first or second language in many different contexts. As a result, there are many varieties of English. Although the great many advances in English automatic speech recognition (ASR) over the past decades, results are usually reported based on test datasets which fail to represent the diversity of English as spoken today around the globe. We present the first release of The Edinburgh International Accents of English Corpus (EdAcc). This dataset attempts to better represent the wide diversity of English, encompassing almost 40 hours of dyadic video call conversations between friends. Unlike other datasets, EdAcc includes a wide range of first and second-language varieties of English and a linguistic background profile of each speaker. Results on latest public, and commercial models show that EdAcc highlights shortcomings of current English ASR models. The best performing model, trained on 680 thousand hours of transcribed data, obtains an average of 19.7% word error rate (WER) -- in contrast to the 2.7% WER obtained when evaluated on US English clean read speech. Across all models, we observe a drop in performance on Indian, Jamaican, and Nigerian English speakers. Recordings, linguistic backgrounds, data statement, and evaluation scripts are released on our website (https://groups.inf.ed.ac.uk/edacc/) under CC-BY-SA license.
new_dataset
0.68491
2303.18118
Camille Garcin
Camille Garcin, Maximilien Servajean, Alexis Joly, Joseph Salmon
A two-head loss function for deep Average-K classification
null
null
null
null
cs.LG cs.CV
http://creativecommons.org/licenses/by-sa/4.0/
Average-K classification is an alternative to top-K classification in which the number of labels returned varies with the ambiguity of the input image but must average to K over all the samples. A simple method to solve this task is to threshold the softmax output of a model trained with the cross-entropy loss. This approach is theoretically proven to be asymptotically consistent, but it is not guaranteed to be optimal for a finite set of samples. In this paper, we propose a new loss function based on a multi-label classification head in addition to the classical softmax. This second head is trained using pseudo-labels generated by thresholding the softmax head while guaranteeing that K classes are returned on average. We show that this approach allows the model to better capture ambiguities between classes and, as a result, to return more consistent sets of possible classes. Experiments on two datasets from the literature demonstrate that our approach outperforms the softmax baseline, as well as several other loss functions more generally designed for weakly supervised multi-label classification. The gains are larger the higher the uncertainty, especially for classes with few samples.
[ { "version": "v1", "created": "Fri, 31 Mar 2023 15:04:53 GMT" } ]
2023-04-03T00:00:00
[ [ "Garcin", "Camille", "" ], [ "Servajean", "Maximilien", "" ], [ "Joly", "Alexis", "" ], [ "Salmon", "Joseph", "" ] ]
TITLE: A two-head loss function for deep Average-K classification ABSTRACT: Average-K classification is an alternative to top-K classification in which the number of labels returned varies with the ambiguity of the input image but must average to K over all the samples. A simple method to solve this task is to threshold the softmax output of a model trained with the cross-entropy loss. This approach is theoretically proven to be asymptotically consistent, but it is not guaranteed to be optimal for a finite set of samples. In this paper, we propose a new loss function based on a multi-label classification head in addition to the classical softmax. This second head is trained using pseudo-labels generated by thresholding the softmax head while guaranteeing that K classes are returned on average. We show that this approach allows the model to better capture ambiguities between classes and, as a result, to return more consistent sets of possible classes. Experiments on two datasets from the literature demonstrate that our approach outperforms the softmax baseline, as well as several other loss functions more generally designed for weakly supervised multi-label classification. The gains are larger the higher the uncertainty, especially for classes with few samples.
no_new_dataset
0.710176
2303.18162
Son T. Luu
Son T. Luu, Khoi Trong Hoang, Tuong Quang Pham, Kiet Van Nguyen, Ngan Luu-Thuy Nguyen
A Multiple Choices Reading Comprehension Corpus for Vietnamese Language Education
null
null
null
null
cs.CL
http://creativecommons.org/licenses/by-nc-nd/4.0/
Machine reading comprehension has been an interesting and challenging task in recent years, with the purpose of extracting useful information from texts. To attain the computer ability to understand the reading text and answer relevant information, we introduce ViMMRC 2.0 - an extension of the previous ViMMRC for the task of multiple-choice reading comprehension in Vietnamese Textbooks which contain the reading articles for students from Grade 1 to Grade 12. This dataset has 699 reading passages which are prose and poems, and 5,273 questions. The questions in the new dataset are not fixed with four options as in the previous version. Moreover, the difficulty of questions is increased, which challenges the models to find the correct choice. The computer must understand the whole context of the reading passage, the question, and the content of each choice to extract the right answers. Hence, we propose the multi-stage approach that combines the multi-step attention network (MAN) with the natural language inference (NLI) task to enhance the performance of the reading comprehension model. Then, we compare the proposed methodology with the baseline BERTology models on the new dataset and the ViMMRC 1.0. Our multi-stage models achieved 58.81% by Accuracy on the test set, which is 5.34% better than the highest BERTology models. From the results of the error analysis, we found the challenge of the reading comprehension models is understanding the implicit context in texts and linking them together in order to find the correct answers. Finally, we hope our new dataset will motivate further research in enhancing the language understanding ability of computers in the Vietnamese language.
[ { "version": "v1", "created": "Fri, 31 Mar 2023 15:54:54 GMT" } ]
2023-04-03T00:00:00
[ [ "Luu", "Son T.", "" ], [ "Hoang", "Khoi Trong", "" ], [ "Pham", "Tuong Quang", "" ], [ "Van Nguyen", "Kiet", "" ], [ "Nguyen", "Ngan Luu-Thuy", "" ] ]
TITLE: A Multiple Choices Reading Comprehension Corpus for Vietnamese Language Education ABSTRACT: Machine reading comprehension has been an interesting and challenging task in recent years, with the purpose of extracting useful information from texts. To attain the computer ability to understand the reading text and answer relevant information, we introduce ViMMRC 2.0 - an extension of the previous ViMMRC for the task of multiple-choice reading comprehension in Vietnamese Textbooks which contain the reading articles for students from Grade 1 to Grade 12. This dataset has 699 reading passages which are prose and poems, and 5,273 questions. The questions in the new dataset are not fixed with four options as in the previous version. Moreover, the difficulty of questions is increased, which challenges the models to find the correct choice. The computer must understand the whole context of the reading passage, the question, and the content of each choice to extract the right answers. Hence, we propose the multi-stage approach that combines the multi-step attention network (MAN) with the natural language inference (NLI) task to enhance the performance of the reading comprehension model. Then, we compare the proposed methodology with the baseline BERTology models on the new dataset and the ViMMRC 1.0. Our multi-stage models achieved 58.81% by Accuracy on the test set, which is 5.34% better than the highest BERTology models. From the results of the error analysis, we found the challenge of the reading comprehension models is understanding the implicit context in texts and linking them together in order to find the correct answers. Finally, we hope our new dataset will motivate further research in enhancing the language understanding ability of computers in the Vietnamese language.
new_dataset
0.718378
2303.18176
Nikhil Churamani
Hatice Gunes and Nikhil Churamani
Affective Computing for Human-Robot Interaction Research: Four Critical Lessons for the Hitchhiker
11 pages, 3 figures, 1 table
null
null
null
cs.RO cs.HC
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
Social Robotics and Human-Robot Interaction (HRI) research relies on different Affective Computing (AC) solutions for sensing, perceiving and understanding human affective behaviour during interactions. This may include utilising off-the-shelf affect perception models that are pre-trained on popular affect recognition benchmarks and directly applied to situated interactions. However, the conditions in situated human-robot interactions differ significantly from the training data and settings of these models. Thus, there is a need to deepen our understanding of how AC solutions can be best leveraged, customised and applied for situated HRI. This paper, while critiquing the existing practices, presents four critical lessons to be noted by the hitchhiker when applying AC for HRI research. These lessons conclude that: (i) The six basic emotions categories are irrelevant in situated interactions, (ii) Affect recognition accuracy (%) improvements are unimportant, (iii) Affect recognition does not generalise across contexts, and (iv) Affect recognition alone is insufficient for adaptation and personalisation. By describing the background and the context for each lesson, and demonstrating how these lessons have been learnt, this paper aims to enable the hitchhiker to successfully and insightfully leverage AC solutions for advancing HRI research.
[ { "version": "v1", "created": "Fri, 31 Mar 2023 16:14:50 GMT" } ]
2023-04-03T00:00:00
[ [ "Gunes", "Hatice", "" ], [ "Churamani", "Nikhil", "" ] ]
TITLE: Affective Computing for Human-Robot Interaction Research: Four Critical Lessons for the Hitchhiker ABSTRACT: Social Robotics and Human-Robot Interaction (HRI) research relies on different Affective Computing (AC) solutions for sensing, perceiving and understanding human affective behaviour during interactions. This may include utilising off-the-shelf affect perception models that are pre-trained on popular affect recognition benchmarks and directly applied to situated interactions. However, the conditions in situated human-robot interactions differ significantly from the training data and settings of these models. Thus, there is a need to deepen our understanding of how AC solutions can be best leveraged, customised and applied for situated HRI. This paper, while critiquing the existing practices, presents four critical lessons to be noted by the hitchhiker when applying AC for HRI research. These lessons conclude that: (i) The six basic emotions categories are irrelevant in situated interactions, (ii) Affect recognition accuracy (%) improvements are unimportant, (iii) Affect recognition does not generalise across contexts, and (iv) Affect recognition alone is insufficient for adaptation and personalisation. By describing the background and the context for each lesson, and demonstrating how these lessons have been learnt, this paper aims to enable the hitchhiker to successfully and insightfully leverage AC solutions for advancing HRI research.
no_new_dataset
0.709636
2303.18178
Jingwei Sun
Jingwei Sun, Zhixu Du, Anna Dai, Saleh Baghersalimi, Alireza Amirshahi, David Atienza, Yiran Chen
Robust and IP-Protecting Vertical Federated Learning against Unexpected Quitting of Parties
null
null
null
null
cs.CR cs.LG
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
Vertical federated learning (VFL) enables a service provider (i.e., active party) who owns labeled features to collaborate with passive parties who possess auxiliary features to improve model performance. Existing VFL approaches, however, have two major vulnerabilities when passive parties unexpectedly quit in the deployment phase of VFL - severe performance degradation and intellectual property (IP) leakage of the active party's labels. In this paper, we propose \textbf{Party-wise Dropout} to improve the VFL model's robustness against the unexpected exit of passive parties and a defense method called \textbf{DIMIP} to protect the active party's IP in the deployment phase. We evaluate our proposed methods on multiple datasets against different inference attacks. The results show that Party-wise Dropout effectively maintains model performance after the passive party quits, and DIMIP successfully disguises label information from the passive party's feature extractor, thereby mitigating IP leakage.
[ { "version": "v1", "created": "Tue, 28 Mar 2023 19:58:28 GMT" } ]
2023-04-03T00:00:00
[ [ "Sun", "Jingwei", "" ], [ "Du", "Zhixu", "" ], [ "Dai", "Anna", "" ], [ "Baghersalimi", "Saleh", "" ], [ "Amirshahi", "Alireza", "" ], [ "Atienza", "David", "" ], [ "Chen", "Yiran", "" ] ]
TITLE: Robust and IP-Protecting Vertical Federated Learning against Unexpected Quitting of Parties ABSTRACT: Vertical federated learning (VFL) enables a service provider (i.e., active party) who owns labeled features to collaborate with passive parties who possess auxiliary features to improve model performance. Existing VFL approaches, however, have two major vulnerabilities when passive parties unexpectedly quit in the deployment phase of VFL - severe performance degradation and intellectual property (IP) leakage of the active party's labels. In this paper, we propose \textbf{Party-wise Dropout} to improve the VFL model's robustness against the unexpected exit of passive parties and a defense method called \textbf{DIMIP} to protect the active party's IP in the deployment phase. We evaluate our proposed methods on multiple datasets against different inference attacks. The results show that Party-wise Dropout effectively maintains model performance after the passive party quits, and DIMIP successfully disguises label information from the passive party's feature extractor, thereby mitigating IP leakage.
no_new_dataset
0.708792
2303.18190
Leon Derczynski
Leon Derczynski, Hannah Rose Kirk, Vidhisha Balachandran, Sachin Kumar, Yulia Tsvetkov, M. R. Leiser, Saif Mohammad
Assessing Language Model Deployment with Risk Cards
null
null
null
null
cs.CL
http://creativecommons.org/licenses/by/4.0/
This paper introduces RiskCards, a framework for structured assessment and documentation of risks associated with an application of language models. As with all language, text generated by language models can be harmful, or used to bring about harm. Automating language generation adds both an element of scale and also more subtle or emergent undesirable tendencies to the generated text. Prior work establishes a wide variety of language model harms to many different actors: existing taxonomies identify categories of harms posed by language models; benchmarks establish automated tests of these harms; and documentation standards for models, tasks and datasets encourage transparent reporting. However, there is no risk-centric framework for documenting the complexity of a landscape in which some risks are shared across models and contexts, while others are specific, and where certain conditions may be required for risks to manifest as harms. RiskCards address this methodological gap by providing a generic framework for assessing the use of a given language model in a given scenario. Each RiskCard makes clear the routes for the risk to manifest harm, their placement in harm taxonomies, and example prompt-output pairs. While RiskCards are designed to be open-source, dynamic and participatory, we present a "starter set" of RiskCards taken from a broad literature survey, each of which details a concrete risk presentation. Language model RiskCards initiate a community knowledge base which permits the mapping of risks and harms to a specific model or its application scenario, ultimately contributing to a better, safer and shared understanding of the risk landscape.
[ { "version": "v1", "created": "Fri, 31 Mar 2023 16:45:42 GMT" } ]
2023-04-03T00:00:00
[ [ "Derczynski", "Leon", "" ], [ "Kirk", "Hannah Rose", "" ], [ "Balachandran", "Vidhisha", "" ], [ "Kumar", "Sachin", "" ], [ "Tsvetkov", "Yulia", "" ], [ "Leiser", "M. R.", "" ], [ "Mohammad", "Saif", "" ] ]
TITLE: Assessing Language Model Deployment with Risk Cards ABSTRACT: This paper introduces RiskCards, a framework for structured assessment and documentation of risks associated with an application of language models. As with all language, text generated by language models can be harmful, or used to bring about harm. Automating language generation adds both an element of scale and also more subtle or emergent undesirable tendencies to the generated text. Prior work establishes a wide variety of language model harms to many different actors: existing taxonomies identify categories of harms posed by language models; benchmarks establish automated tests of these harms; and documentation standards for models, tasks and datasets encourage transparent reporting. However, there is no risk-centric framework for documenting the complexity of a landscape in which some risks are shared across models and contexts, while others are specific, and where certain conditions may be required for risks to manifest as harms. RiskCards address this methodological gap by providing a generic framework for assessing the use of a given language model in a given scenario. Each RiskCard makes clear the routes for the risk to manifest harm, their placement in harm taxonomies, and example prompt-output pairs. While RiskCards are designed to be open-source, dynamic and participatory, we present a "starter set" of RiskCards taken from a broad literature survey, each of which details a concrete risk presentation. Language model RiskCards initiate a community knowledge base which permits the mapping of risks and harms to a specific model or its application scenario, ultimately contributing to a better, safer and shared understanding of the risk landscape.
no_new_dataset
0.708635
2303.18191
Xiaogeng Liu
Xiaogeng Liu, Minghui Li, Haoyu Wang, Shengshan Hu, Dengpan Ye, Hai Jin, Libing Wu, Chaowei Xiao
Detecting Backdoors During the Inference Stage Based on Corruption Robustness Consistency
Accepted by CVPR2023. Code is available at https://github.com/CGCL-codes/TeCo
null
null
null
cs.CR cs.AI cs.LG
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
Deep neural networks are proven to be vulnerable to backdoor attacks. Detecting the trigger samples during the inference stage, i.e., the test-time trigger sample detection, can prevent the backdoor from being triggered. However, existing detection methods often require the defenders to have high accessibility to victim models, extra clean data, or knowledge about the appearance of backdoor triggers, limiting their practicality. In this paper, we propose the test-time corruption robustness consistency evaluation (TeCo), a novel test-time trigger sample detection method that only needs the hard-label outputs of the victim models without any extra information. Our journey begins with the intriguing observation that the backdoor-infected models have similar performance across different image corruptions for the clean images, but perform discrepantly for the trigger samples. Based on this phenomenon, we design TeCo to evaluate test-time robustness consistency by calculating the deviation of severity that leads to predictions' transition across different corruptions. Extensive experiments demonstrate that compared with state-of-the-art defenses, which even require either certain information about the trigger types or accessibility of clean data, TeCo outperforms them on different backdoor attacks, datasets, and model architectures, enjoying a higher AUROC by 10% and 5 times of stability.
[ { "version": "v1", "created": "Mon, 27 Mar 2023 07:10:37 GMT" } ]
2023-04-03T00:00:00
[ [ "Liu", "Xiaogeng", "" ], [ "Li", "Minghui", "" ], [ "Wang", "Haoyu", "" ], [ "Hu", "Shengshan", "" ], [ "Ye", "Dengpan", "" ], [ "Jin", "Hai", "" ], [ "Wu", "Libing", "" ], [ "Xiao", "Chaowei", "" ] ]
TITLE: Detecting Backdoors During the Inference Stage Based on Corruption Robustness Consistency ABSTRACT: Deep neural networks are proven to be vulnerable to backdoor attacks. Detecting the trigger samples during the inference stage, i.e., the test-time trigger sample detection, can prevent the backdoor from being triggered. However, existing detection methods often require the defenders to have high accessibility to victim models, extra clean data, or knowledge about the appearance of backdoor triggers, limiting their practicality. In this paper, we propose the test-time corruption robustness consistency evaluation (TeCo), a novel test-time trigger sample detection method that only needs the hard-label outputs of the victim models without any extra information. Our journey begins with the intriguing observation that the backdoor-infected models have similar performance across different image corruptions for the clean images, but perform discrepantly for the trigger samples. Based on this phenomenon, we design TeCo to evaluate test-time robustness consistency by calculating the deviation of severity that leads to predictions' transition across different corruptions. Extensive experiments demonstrate that compared with state-of-the-art defenses, which even require either certain information about the trigger types or accessibility of clean data, TeCo outperforms them on different backdoor attacks, datasets, and model architectures, enjoying a higher AUROC by 10% and 5 times of stability.
no_new_dataset
0.707582
2303.18193
Mallikarjun Byrasandra Ramalinga Reddy
Mallikarjun B R, Xingang Pan, Mohamed Elgharib, Christian Theobalt
GVP: Generative Volumetric Primitives
https://vcai.mpi-inf.mpg.de/projects/GVP/index.html
null
null
null
cs.CV cs.GR cs.LG
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
Advances in 3D-aware generative models have pushed the boundary of image synthesis with explicit camera control. To achieve high-resolution image synthesis, several attempts have been made to design efficient generators, such as hybrid architectures with both 3D and 2D components. However, such a design compromises multiview consistency, and the design of a pure 3D generator with high resolution is still an open problem. In this work, we present Generative Volumetric Primitives (GVP), the first pure 3D generative model that can sample and render 512-resolution images in real-time. GVP jointly models a number of volumetric primitives and their spatial information, both of which can be efficiently generated via a 2D convolutional network. The mixture of these primitives naturally captures the sparsity and correspondence in the 3D volume. The training of such a generator with a high degree of freedom is made possible through a knowledge distillation technique. Experiments on several datasets demonstrate superior efficiency and 3D consistency of GVP over the state-of-the-art.
[ { "version": "v1", "created": "Fri, 31 Mar 2023 16:50:23 GMT" } ]
2023-04-03T00:00:00
[ [ "R", "Mallikarjun B", "" ], [ "Pan", "Xingang", "" ], [ "Elgharib", "Mohamed", "" ], [ "Theobalt", "Christian", "" ] ]
TITLE: GVP: Generative Volumetric Primitives ABSTRACT: Advances in 3D-aware generative models have pushed the boundary of image synthesis with explicit camera control. To achieve high-resolution image synthesis, several attempts have been made to design efficient generators, such as hybrid architectures with both 3D and 2D components. However, such a design compromises multiview consistency, and the design of a pure 3D generator with high resolution is still an open problem. In this work, we present Generative Volumetric Primitives (GVP), the first pure 3D generative model that can sample and render 512-resolution images in real-time. GVP jointly models a number of volumetric primitives and their spatial information, both of which can be efficiently generated via a 2D convolutional network. The mixture of these primitives naturally captures the sparsity and correspondence in the 3D volume. The training of such a generator with a high degree of freedom is made possible through a knowledge distillation technique. Experiments on several datasets demonstrate superior efficiency and 3D consistency of GVP over the state-of-the-art.
no_new_dataset
0.708377
2303.18210
Chuangguan Ye
Chuangguan Ye, Hongyuan Zhu, Bo Zhang, Tao Chen
A Closer Look at Few-Shot 3D Point Cloud Classification
Accepted by IJCV 2023
null
10.1007/s11263-022-01731-4
null
cs.CV
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
In recent years, research on few-shot learning (FSL) has been fast-growing in the 2D image domain due to the less requirement for labeled training data and greater generalization for novel classes. However, its application in 3D point cloud data is relatively under-explored. Not only need to distinguish unseen classes as in the 2D domain, 3D FSL is more challenging in terms of irregular structures, subtle inter-class differences, and high intra-class variances {when trained on a low number of data.} Moreover, different architectures and learning algorithms make it difficult to study the effectiveness of existing 2D FSL algorithms when migrating to the 3D domain. In this work, for the first time, we perform systematic and extensive investigations of directly applying recent 2D FSL works to 3D point cloud related backbone networks and thus suggest a strong learning baseline for few-shot 3D point cloud classification. Furthermore, we propose a new network, Point-cloud Correlation Interaction (PCIA), with three novel plug-and-play components called Salient-Part Fusion (SPF) module, Self-Channel Interaction Plus (SCI+) module, and Cross-Instance Fusion Plus (CIF+) module to obtain more representative embeddings and improve the feature distinction. These modules can be inserted into most FSL algorithms with minor changes and significantly improve the performance. Experimental results on three benchmark datasets, ModelNet40-FS, ShapeNet70-FS, and ScanObjectNN-FS, demonstrate that our method achieves state-of-the-art performance for the 3D FSL task. Code and datasets are available at https://github.com/cgye96/A_Closer_Look_At_3DFSL.
[ { "version": "v1", "created": "Fri, 31 Mar 2023 17:01:13 GMT" } ]
2023-04-03T00:00:00
[ [ "Ye", "Chuangguan", "" ], [ "Zhu", "Hongyuan", "" ], [ "Zhang", "Bo", "" ], [ "Chen", "Tao", "" ] ]
TITLE: A Closer Look at Few-Shot 3D Point Cloud Classification ABSTRACT: In recent years, research on few-shot learning (FSL) has been fast-growing in the 2D image domain due to the less requirement for labeled training data and greater generalization for novel classes. However, its application in 3D point cloud data is relatively under-explored. Not only need to distinguish unseen classes as in the 2D domain, 3D FSL is more challenging in terms of irregular structures, subtle inter-class differences, and high intra-class variances {when trained on a low number of data.} Moreover, different architectures and learning algorithms make it difficult to study the effectiveness of existing 2D FSL algorithms when migrating to the 3D domain. In this work, for the first time, we perform systematic and extensive investigations of directly applying recent 2D FSL works to 3D point cloud related backbone networks and thus suggest a strong learning baseline for few-shot 3D point cloud classification. Furthermore, we propose a new network, Point-cloud Correlation Interaction (PCIA), with three novel plug-and-play components called Salient-Part Fusion (SPF) module, Self-Channel Interaction Plus (SCI+) module, and Cross-Instance Fusion Plus (CIF+) module to obtain more representative embeddings and improve the feature distinction. These modules can be inserted into most FSL algorithms with minor changes and significantly improve the performance. Experimental results on three benchmark datasets, ModelNet40-FS, ShapeNet70-FS, and ScanObjectNN-FS, demonstrate that our method achieves state-of-the-art performance for the 3D FSL task. Code and datasets are available at https://github.com/cgye96/A_Closer_Look_At_3DFSL.
no_new_dataset
0.709617
2303.18219
Shan Lin
Shan Lin, Yuheng Zhi, and Michael C. Yip
SemHint-MD: Learning from Noisy Semantic Labels for Self-Supervised Monocular Depth Estimation
null
null
null
null
cs.CV
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
Without ground truth supervision, self-supervised depth estimation can be trapped in a local minimum due to the gradient-locality issue of the photometric loss. In this paper, we present a framework to enhance depth by leveraging semantic segmentation to guide the network to jump out of the local minimum. Prior works have proposed to share encoders between these two tasks or explicitly align them based on priors like the consistency between edges in the depth and segmentation maps. Yet, these methods usually require ground truth or high-quality pseudo labels, which may not be easily accessible in real-world applications. In contrast, we investigate self-supervised depth estimation along with a segmentation branch that is supervised with noisy labels provided by models pre-trained with limited data. We extend parameter sharing from the encoder to the decoder and study the influence of different numbers of shared decoder parameters on model performance. Also, we propose to use cross-task information to refine current depth and segmentation predictions to generate pseudo-depth and semantic labels for training. The advantages of the proposed method are demonstrated through extensive experiments on the KITTI benchmark and a downstream task for endoscopic tissue deformation tracking.
[ { "version": "v1", "created": "Fri, 31 Mar 2023 17:20:27 GMT" } ]
2023-04-03T00:00:00
[ [ "Lin", "Shan", "" ], [ "Zhi", "Yuheng", "" ], [ "Yip", "Michael C.", "" ] ]
TITLE: SemHint-MD: Learning from Noisy Semantic Labels for Self-Supervised Monocular Depth Estimation ABSTRACT: Without ground truth supervision, self-supervised depth estimation can be trapped in a local minimum due to the gradient-locality issue of the photometric loss. In this paper, we present a framework to enhance depth by leveraging semantic segmentation to guide the network to jump out of the local minimum. Prior works have proposed to share encoders between these two tasks or explicitly align them based on priors like the consistency between edges in the depth and segmentation maps. Yet, these methods usually require ground truth or high-quality pseudo labels, which may not be easily accessible in real-world applications. In contrast, we investigate self-supervised depth estimation along with a segmentation branch that is supervised with noisy labels provided by models pre-trained with limited data. We extend parameter sharing from the encoder to the decoder and study the influence of different numbers of shared decoder parameters on model performance. Also, we propose to use cross-task information to refine current depth and segmentation predictions to generate pseudo-depth and semantic labels for training. The advantages of the proposed method are demonstrated through extensive experiments on the KITTI benchmark and a downstream task for endoscopic tissue deformation tracking.
no_new_dataset
0.710245
2303.18247
Xiao Zhou
Xiao Zhou, Yujie Zhong, Zhen Cheng, Fan Liang, Lin Ma
Adaptive Sparse Pairwise Loss for Object Re-Identification
Accepted by CVPR 2023
null
null
null
cs.CV
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
Object re-identification (ReID) aims to find instances with the same identity as the given probe from a large gallery. Pairwise losses play an important role in training a strong ReID network. Existing pairwise losses densely exploit each instance as an anchor and sample its triplets in a mini-batch. This dense sampling mechanism inevitably introduces positive pairs that share few visual similarities, which can be harmful to the training. To address this problem, we propose a novel loss paradigm termed Sparse Pairwise (SP) loss that only leverages few appropriate pairs for each class in a mini-batch, and empirically demonstrate that it is sufficient for the ReID tasks. Based on the proposed loss framework, we propose an adaptive positive mining strategy that can dynamically adapt to diverse intra-class variations. Extensive experiments show that SP loss and its adaptive variant AdaSP loss outperform other pairwise losses, and achieve state-of-the-art performance across several ReID benchmarks. Code is available at https://github.com/Astaxanthin/AdaSP.
[ { "version": "v1", "created": "Fri, 31 Mar 2023 17:59:44 GMT" } ]
2023-04-03T00:00:00
[ [ "Zhou", "Xiao", "" ], [ "Zhong", "Yujie", "" ], [ "Cheng", "Zhen", "" ], [ "Liang", "Fan", "" ], [ "Ma", "Lin", "" ] ]
TITLE: Adaptive Sparse Pairwise Loss for Object Re-Identification ABSTRACT: Object re-identification (ReID) aims to find instances with the same identity as the given probe from a large gallery. Pairwise losses play an important role in training a strong ReID network. Existing pairwise losses densely exploit each instance as an anchor and sample its triplets in a mini-batch. This dense sampling mechanism inevitably introduces positive pairs that share few visual similarities, which can be harmful to the training. To address this problem, we propose a novel loss paradigm termed Sparse Pairwise (SP) loss that only leverages few appropriate pairs for each class in a mini-batch, and empirically demonstrate that it is sufficient for the ReID tasks. Based on the proposed loss framework, we propose an adaptive positive mining strategy that can dynamically adapt to diverse intra-class variations. Extensive experiments show that SP loss and its adaptive variant AdaSP loss outperform other pairwise losses, and achieve state-of-the-art performance across several ReID benchmarks. Code is available at https://github.com/Astaxanthin/AdaSP.
no_new_dataset
0.709629
1701.03017
Stefano Cresci
Stefano Cresci, Roberto Di Pietro, Marinella Petrocchi, Angelo Spognardi, Maurizio Tesconi
The paradigm-shift of social spambots: Evidence, theories, and tools for the arms race
Post-print of the article published in the Proceedings of 26th WWW, 2017, Companion Volume (Web Science Track, Perth, Australia, 3-7 April, 2017)
Proceedings of the 26th International Conference on World Wide Web Companion, 2017
10.1145/3041021.3055135
null
cs.SI
http://creativecommons.org/licenses/by-nc-nd/4.0/
Recent studies in social media spam and automation provide anecdotal argumentation of the rise of a new generation of spambots, so-called social spambots. Here, for the first time, we extensively study this novel phenomenon on Twitter and we provide quantitative evidence that a paradigm-shift exists in spambot design. First, we measure current Twitter's capabilities of detecting the new social spambots. Later, we assess the human performance in discriminating between genuine accounts, social spambots, and traditional spambots. Then, we benchmark several state-of-the-art techniques proposed by the academic literature. Results show that neither Twitter, nor humans, nor cutting-edge applications are currently capable of accurately detecting the new social spambots. Our results call for new approaches capable of turning the tide in the fight against this raising phenomenon. We conclude by reviewing the latest literature on spambots detection and we highlight an emerging common research trend based on the analysis of collective behaviors. Insights derived from both our extensive experimental campaign and survey shed light on the most promising directions of research and lay the foundations for the arms race against the novel social spambots. Finally, to foster research on this novel phenomenon, we make publicly available to the scientific community all the datasets used in this study.
[ { "version": "v1", "created": "Wed, 11 Jan 2017 15:33:15 GMT" }, { "version": "v2", "created": "Thu, 30 Mar 2023 10:38:23 GMT" } ]
2023-03-31T00:00:00
[ [ "Cresci", "Stefano", "" ], [ "Di Pietro", "Roberto", "" ], [ "Petrocchi", "Marinella", "" ], [ "Spognardi", "Angelo", "" ], [ "Tesconi", "Maurizio", "" ] ]
TITLE: The paradigm-shift of social spambots: Evidence, theories, and tools for the arms race ABSTRACT: Recent studies in social media spam and automation provide anecdotal argumentation of the rise of a new generation of spambots, so-called social spambots. Here, for the first time, we extensively study this novel phenomenon on Twitter and we provide quantitative evidence that a paradigm-shift exists in spambot design. First, we measure current Twitter's capabilities of detecting the new social spambots. Later, we assess the human performance in discriminating between genuine accounts, social spambots, and traditional spambots. Then, we benchmark several state-of-the-art techniques proposed by the academic literature. Results show that neither Twitter, nor humans, nor cutting-edge applications are currently capable of accurately detecting the new social spambots. Our results call for new approaches capable of turning the tide in the fight against this raising phenomenon. We conclude by reviewing the latest literature on spambots detection and we highlight an emerging common research trend based on the analysis of collective behaviors. Insights derived from both our extensive experimental campaign and survey shed light on the most promising directions of research and lay the foundations for the arms race against the novel social spambots. Finally, to foster research on this novel phenomenon, we make publicly available to the scientific community all the datasets used in this study.
no_new_dataset
0.701879
2002.10243
Tianyu Cui
Tianyu Cui, Aki Havulinna, Pekka Marttinen, Samuel Kaski
Informative Bayesian Neural Network Priors for Weak Signals
25 pages, 8 figures, 4 tables
null
10.1214/21-BA1291
null
stat.ML cs.LG
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
Encoding domain knowledge into the prior over the high-dimensional weight space of a neural network is challenging but essential in applications with limited data and weak signals. Two types of domain knowledge are commonly available in scientific applications: 1. feature sparsity (fraction of features deemed relevant); 2. signal-to-noise ratio, quantified, for instance, as the proportion of variance explained (PVE). We show how to encode both types of domain knowledge into the widely used Gaussian scale mixture priors with Automatic Relevance Determination. Specifically, we propose a new joint prior over the local (i.e., feature-specific) scale parameters that encodes knowledge about feature sparsity, and a Stein gradient optimization to tune the hyperparameters in such a way that the distribution induced on the model's PVE matches the prior distribution. We show empirically that the new prior improves prediction accuracy, compared to existing neural network priors, on several publicly available datasets and in a genetics application where signals are weak and sparse, often outperforming even computationally intensive cross-validation for hyperparameter tuning.
[ { "version": "v1", "created": "Mon, 24 Feb 2020 13:43:44 GMT" }, { "version": "v2", "created": "Thu, 7 Jan 2021 14:55:29 GMT" } ]
2023-03-31T00:00:00
[ [ "Cui", "Tianyu", "" ], [ "Havulinna", "Aki", "" ], [ "Marttinen", "Pekka", "" ], [ "Kaski", "Samuel", "" ] ]
TITLE: Informative Bayesian Neural Network Priors for Weak Signals ABSTRACT: Encoding domain knowledge into the prior over the high-dimensional weight space of a neural network is challenging but essential in applications with limited data and weak signals. Two types of domain knowledge are commonly available in scientific applications: 1. feature sparsity (fraction of features deemed relevant); 2. signal-to-noise ratio, quantified, for instance, as the proportion of variance explained (PVE). We show how to encode both types of domain knowledge into the widely used Gaussian scale mixture priors with Automatic Relevance Determination. Specifically, we propose a new joint prior over the local (i.e., feature-specific) scale parameters that encodes knowledge about feature sparsity, and a Stein gradient optimization to tune the hyperparameters in such a way that the distribution induced on the model's PVE matches the prior distribution. We show empirically that the new prior improves prediction accuracy, compared to existing neural network priors, on several publicly available datasets and in a genetics application where signals are weak and sparse, often outperforming even computationally intensive cross-validation for hyperparameter tuning.
no_new_dataset
0.710038
2003.06777
Chi Zhang
Chi Zhang, Yujun Cai, Guosheng Lin, Chunhua Shen
DeepEMD: Differentiable Earth Mover's Distance for Few-Shot Learning
DeepEMD V2, accepted by TPAMI
null
null
null
cs.CV cs.LG eess.IV
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
In this work, we develop methods for few-shot image classification from a new perspective of optimal matching between image regions. We employ the Earth Mover's Distance (EMD) as a metric to compute a structural distance between dense image representations to determine image relevance. The EMD generates the optimal matching flows between structural elements that have the minimum matching cost, which is used to calculate the image distance for classification. To generate the important weights of elements in the EMD formulation, we design a cross-reference mechanism, which can effectively alleviate the adverse impact caused by the cluttered background and large intra-class appearance variations. To implement k-shot classification, we propose to learn a structured fully connected layer that can directly classify dense image representations with the EMD. Based on the implicit function theorem, the EMD can be inserted as a layer into the network for end-to-end training. Our extensive experiments validate the effectiveness of our algorithm which outperforms state-of-the-art methods by a significant margin on five widely used few-shot classification benchmarks, namely, miniImageNet, tieredImageNet, Fewshot-CIFAR100 (FC100), Caltech-UCSD Birds-200-2011 (CUB), and CIFAR-FewShot (CIFAR-FS). We also demonstrate the effectiveness of our method on the image retrieval task in our experiments.
[ { "version": "v1", "created": "Sun, 15 Mar 2020 08:13:16 GMT" }, { "version": "v2", "created": "Wed, 25 Mar 2020 12:56:06 GMT" }, { "version": "v3", "created": "Fri, 8 May 2020 02:27:17 GMT" }, { "version": "v4", "created": "Sat, 29 Jan 2022 11:34:11 GMT" }, { "version": "v5", "created": "Thu, 30 Mar 2023 10:48:54 GMT" } ]
2023-03-31T00:00:00
[ [ "Zhang", "Chi", "" ], [ "Cai", "Yujun", "" ], [ "Lin", "Guosheng", "" ], [ "Shen", "Chunhua", "" ] ]
TITLE: DeepEMD: Differentiable Earth Mover's Distance for Few-Shot Learning ABSTRACT: In this work, we develop methods for few-shot image classification from a new perspective of optimal matching between image regions. We employ the Earth Mover's Distance (EMD) as a metric to compute a structural distance between dense image representations to determine image relevance. The EMD generates the optimal matching flows between structural elements that have the minimum matching cost, which is used to calculate the image distance for classification. To generate the important weights of elements in the EMD formulation, we design a cross-reference mechanism, which can effectively alleviate the adverse impact caused by the cluttered background and large intra-class appearance variations. To implement k-shot classification, we propose to learn a structured fully connected layer that can directly classify dense image representations with the EMD. Based on the implicit function theorem, the EMD can be inserted as a layer into the network for end-to-end training. Our extensive experiments validate the effectiveness of our algorithm which outperforms state-of-the-art methods by a significant margin on five widely used few-shot classification benchmarks, namely, miniImageNet, tieredImageNet, Fewshot-CIFAR100 (FC100), Caltech-UCSD Birds-200-2011 (CUB), and CIFAR-FewShot (CIFAR-FS). We also demonstrate the effectiveness of our method on the image retrieval task in our experiments.
no_new_dataset
0.710025
2009.00826
Renchi Yang Dr.
Renchi Yang, Jieming Shi, Xiaokui Xiao, Yin Yang, Sourav S. Bhowmick, Juncheng Liu
PANE: scalable and effective attributed network embedding
Paper published in VLDBJ 2023. Extended version of PVLDB Vol. 14 (VLDB 2021) paper titled "Scaling Attributed Network Embedding to Massive Graphs"
null
10.1007/s00778-023-00790-4
null
cs.SI
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
Given a graph G where each node is associated with a set of attributes, attributed network embedding (ANE) maps each node v in G to a compact vector Xv, which can be used in downstream machine learning tasks. Ideally, Xv should capture node v's affinity to each attribute, which considers not only v's own attribute associations, but also those of its connected nodes along edges in G. It is challenging to obtain high-utility embeddings that enable accurate predictions; scaling effective ANE computation to massive graphs pushes the difficulty of the problem to a whole new level. Existing solutions largely fail on such graphs, leading to prohibitive costs, low-quality embeddings, or both. This paper proposes PANE, an effective and scalable approach to ANE computation for massive graphs that achieves state-of-the-art result quality on multiple benchmark datasets. PANE obtains high scalability and effectiveness through 3 main algorithmic designs. First, it formulates the learning objective based on a novel random walk model for attributed networks. Second, PANE includes a highly efficient solver for the above optimization problem, whose key module is a carefully designed initialization of the embeddings, which drastically reduces the number of iterations required to converge. Finally, PANE utilizes multi-core CPUs through non-trivial parallelization of the above solver, which achieves scalability while retaining the high quality of the resulting embeddings. The performance of PANE depends upon the number of attributes in the input network. To handle large networks with numerous attributes, we further extend PANE to PANE++. Extensive experiments, comparing 10 existing approaches on 8 real datasets, demonstrate that PANE and PANE++ consistently outperform all existing methods in terms of result quality, while being orders of magnitude faster.
[ { "version": "v1", "created": "Wed, 2 Sep 2020 05:15:03 GMT" }, { "version": "v2", "created": "Thu, 30 Mar 2023 08:49:45 GMT" } ]
2023-03-31T00:00:00
[ [ "Yang", "Renchi", "" ], [ "Shi", "Jieming", "" ], [ "Xiao", "Xiaokui", "" ], [ "Yang", "Yin", "" ], [ "Bhowmick", "Sourav S.", "" ], [ "Liu", "Juncheng", "" ] ]
TITLE: PANE: scalable and effective attributed network embedding ABSTRACT: Given a graph G where each node is associated with a set of attributes, attributed network embedding (ANE) maps each node v in G to a compact vector Xv, which can be used in downstream machine learning tasks. Ideally, Xv should capture node v's affinity to each attribute, which considers not only v's own attribute associations, but also those of its connected nodes along edges in G. It is challenging to obtain high-utility embeddings that enable accurate predictions; scaling effective ANE computation to massive graphs pushes the difficulty of the problem to a whole new level. Existing solutions largely fail on such graphs, leading to prohibitive costs, low-quality embeddings, or both. This paper proposes PANE, an effective and scalable approach to ANE computation for massive graphs that achieves state-of-the-art result quality on multiple benchmark datasets. PANE obtains high scalability and effectiveness through 3 main algorithmic designs. First, it formulates the learning objective based on a novel random walk model for attributed networks. Second, PANE includes a highly efficient solver for the above optimization problem, whose key module is a carefully designed initialization of the embeddings, which drastically reduces the number of iterations required to converge. Finally, PANE utilizes multi-core CPUs through non-trivial parallelization of the above solver, which achieves scalability while retaining the high quality of the resulting embeddings. The performance of PANE depends upon the number of attributes in the input network. To handle large networks with numerous attributes, we further extend PANE to PANE++. Extensive experiments, comparing 10 existing approaches on 8 real datasets, demonstrate that PANE and PANE++ consistently outperform all existing methods in terms of result quality, while being orders of magnitude faster.
no_new_dataset
0.70844
2009.11917
Benson Tsz Kin Leung
Benson Tsz Kin Leung
Learning in a Small/Big World
null
null
null
null
econ.TH cs.AI econ.GN q-fin.EC
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
Complexity and limited ability have profound effect on how we learn and make decisions under uncertainty. Using the theory of finite automaton to model belief formation, this paper studies the characteristics of optimal learning behavior in small and big worlds, where the complexity of the environment is low and high, respectively, relative to the cognitive ability of the decision maker. Optimal behavior is well approximated by the Bayesian benchmark in very small world but is more different as the world gets bigger. In addition, in big worlds, the optimal learning behavior could exhibit a wide range of well-documented non-Bayesian learning behavior, including the use of heuristics, correlation neglect, persistent over-confidence, inattentive learning, and other behaviors of model simplification or misspecification. These results establish a clear and testable relationship among the prominence of non-Bayesian learning behavior, complexity, and cognitive ability.
[ { "version": "v1", "created": "Thu, 24 Sep 2020 19:25:02 GMT" }, { "version": "v2", "created": "Mon, 28 Sep 2020 09:53:05 GMT" }, { "version": "v3", "created": "Thu, 8 Oct 2020 19:42:13 GMT" }, { "version": "v4", "created": "Tue, 13 Oct 2020 14:25:01 GMT" }, { "version": "v5", "created": "Sat, 17 Oct 2020 20:51:54 GMT" }, { "version": "v6", "created": "Sun, 1 Nov 2020 16:31:05 GMT" }, { "version": "v7", "created": "Fri, 13 Nov 2020 17:45:08 GMT" }, { "version": "v8", "created": "Thu, 30 Mar 2023 09:38:30 GMT" } ]
2023-03-31T00:00:00
[ [ "Leung", "Benson Tsz Kin", "" ] ]
TITLE: Learning in a Small/Big World ABSTRACT: Complexity and limited ability have profound effect on how we learn and make decisions under uncertainty. Using the theory of finite automaton to model belief formation, this paper studies the characteristics of optimal learning behavior in small and big worlds, where the complexity of the environment is low and high, respectively, relative to the cognitive ability of the decision maker. Optimal behavior is well approximated by the Bayesian benchmark in very small world but is more different as the world gets bigger. In addition, in big worlds, the optimal learning behavior could exhibit a wide range of well-documented non-Bayesian learning behavior, including the use of heuristics, correlation neglect, persistent over-confidence, inattentive learning, and other behaviors of model simplification or misspecification. These results establish a clear and testable relationship among the prominence of non-Bayesian learning behavior, complexity, and cognitive ability.
no_new_dataset
0.713076
2103.07908
Pou-Chun Kung
Pou-Chun Kung and Chieh-Chih Wang and Wen-Chieh Lin
A Normal Distribution Transform-Based Radar Odometry Designed For Scanning and Automotive Radars
Accepted for publication in ICRA 2021. Code is available: For scanning RO, see https://github.com/kungfrank/pw_ndt_radar_scan_matching . For automotive RO, see https://github.com/kungfrank/pw_ndt_automotive_radar_scan_matching
null
null
null
cs.RO
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
Existing radar sensors can be classified into automotive and scanning radars. While most radar odometry (RO) methods are only designed for a specific type of radar, our RO method adapts to both scanning and automotive radars. Our RO is simple yet effective, where the pipeline consists of thresholding, probabilistic submap building, and an NDT-based radar scan matching. The proposed RO has been tested on two public radar datasets: the Oxford Radar RobotCar dataset and the nuScenes dataset, which provide scanning and automotive radar data respectively. The results show that our approach surpasses state-of-the-art RO using either automotive or scanning radar by reducing translational error by 51% and 30%, respectively, and rotational error by 17% and 29%, respectively. Besides, we show that our RO achieves centimeter-level accuracy as lidar odometry, and automotive and scanning RO have similar accuracy.
[ { "version": "v1", "created": "Sun, 14 Mar 2021 12:22:32 GMT" }, { "version": "v2", "created": "Tue, 16 Mar 2021 14:56:48 GMT" }, { "version": "v3", "created": "Thu, 30 Mar 2023 05:52:46 GMT" } ]
2023-03-31T00:00:00
[ [ "Kung", "Pou-Chun", "" ], [ "Wang", "Chieh-Chih", "" ], [ "Lin", "Wen-Chieh", "" ] ]
TITLE: A Normal Distribution Transform-Based Radar Odometry Designed For Scanning and Automotive Radars ABSTRACT: Existing radar sensors can be classified into automotive and scanning radars. While most radar odometry (RO) methods are only designed for a specific type of radar, our RO method adapts to both scanning and automotive radars. Our RO is simple yet effective, where the pipeline consists of thresholding, probabilistic submap building, and an NDT-based radar scan matching. The proposed RO has been tested on two public radar datasets: the Oxford Radar RobotCar dataset and the nuScenes dataset, which provide scanning and automotive radar data respectively. The results show that our approach surpasses state-of-the-art RO using either automotive or scanning radar by reducing translational error by 51% and 30%, respectively, and rotational error by 17% and 29%, respectively. Besides, we show that our RO achieves centimeter-level accuracy as lidar odometry, and automotive and scanning RO have similar accuracy.
no_new_dataset
0.703982
2109.05765
Kaichen Zhou
Kaichen Zhou, Lanqing Hong, Shoukang Hu, Fengwei Zhou, Binxin Ru, Jiashi Feng, Zhenguo Li
DHA: End-to-End Joint Optimization of Data Augmentation Policy, Hyper-parameter and Architecture
null
Transactions on Machine Learning Research 2022
null
null
cs.LG cs.CV
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
Automated machine learning (AutoML) usually involves several crucial components, such as Data Augmentation (DA) policy, Hyper-Parameter Optimization (HPO), and Neural Architecture Search (NAS). Although many strategies have been developed for automating these components in separation, joint optimization of these components remains challenging due to the largely increased search dimension and the variant input types of each component. In parallel to this, the common practice of searching for the optimal architecture first and then retraining it before deployment in NAS often suffers from low performance correlation between the searching and retraining stages. An end-to-end solution that integrates the AutoML components and returns a ready-to-use model at the end of the search is desirable. In view of these, we propose DHA, which achieves joint optimization of Data augmentation policy, Hyper-parameter and Architecture. Specifically, end-to-end NAS is achieved in a differentiable manner by optimizing a compressed lower-dimensional feature space, while DA policy and HPO are regarded as dynamic schedulers, which adapt themselves to the update of network parameters and network architecture at the same time. Experiments show that DHA achieves state-of-the-art (SOTA) results on various datasets and search spaces. To the best of our knowledge, we are the first to efficiently and jointly optimize DA policy, NAS, and HPO in an end-to-end manner without retraining.
[ { "version": "v1", "created": "Mon, 13 Sep 2021 08:12:50 GMT" }, { "version": "v2", "created": "Thu, 3 Nov 2022 12:33:19 GMT" } ]
2023-03-31T00:00:00
[ [ "Zhou", "Kaichen", "" ], [ "Hong", "Lanqing", "" ], [ "Hu", "Shoukang", "" ], [ "Zhou", "Fengwei", "" ], [ "Ru", "Binxin", "" ], [ "Feng", "Jiashi", "" ], [ "Li", "Zhenguo", "" ] ]
TITLE: DHA: End-to-End Joint Optimization of Data Augmentation Policy, Hyper-parameter and Architecture ABSTRACT: Automated machine learning (AutoML) usually involves several crucial components, such as Data Augmentation (DA) policy, Hyper-Parameter Optimization (HPO), and Neural Architecture Search (NAS). Although many strategies have been developed for automating these components in separation, joint optimization of these components remains challenging due to the largely increased search dimension and the variant input types of each component. In parallel to this, the common practice of searching for the optimal architecture first and then retraining it before deployment in NAS often suffers from low performance correlation between the searching and retraining stages. An end-to-end solution that integrates the AutoML components and returns a ready-to-use model at the end of the search is desirable. In view of these, we propose DHA, which achieves joint optimization of Data augmentation policy, Hyper-parameter and Architecture. Specifically, end-to-end NAS is achieved in a differentiable manner by optimizing a compressed lower-dimensional feature space, while DA policy and HPO are regarded as dynamic schedulers, which adapt themselves to the update of network parameters and network architecture at the same time. Experiments show that DHA achieves state-of-the-art (SOTA) results on various datasets and search spaces. To the best of our knowledge, we are the first to efficiently and jointly optimize DA policy, NAS, and HPO in an end-to-end manner without retraining.
no_new_dataset
0.709831
2112.13747
Xiaofeng Pan
Xiaofeng Pan, Yibin Shen, Jing Zhang, Xu He, Yang Huang, Hong Wen, Chengjun Mao and Bo Cao
MOEF: Modeling Occasion Evolution in Frequency Domain for Promotion-Aware Click-Through Rate Prediction
null
null
null
null
cs.LG cs.IR
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
Promotions are becoming more important and prevalent in e-commerce to attract customers and boost sales, leading to frequent changes of occasions, which drives users to behave differently. In such situations, most existing Click-Through Rate (CTR) models can't generalize well to online serving due to distribution uncertainty of the upcoming occasion. In this paper, we propose a novel CTR model named MOEF for recommendations under frequent changes of occasions. Firstly, we design a time series that consists of occasion signals generated from the online business scenario. Since occasion signals are more discriminative in the frequency domain, we apply Fourier Transformation to sliding time windows upon the time series, obtaining a sequence of frequency spectrum which is then processed by Occasion Evolution Layer (OEL). In this way, a high-order occasion representation can be learned to handle the online distribution uncertainty. Moreover, we adopt multiple experts to learn feature representations from multiple aspects, which are guided by the occasion representation via an attention mechanism. Accordingly, a mixture of feature representations is obtained adaptively for different occasions to predict the final CTR. Experimental results on real-world datasets validate the superiority of MOEF and online A/B tests also show MOEF outperforms representative CTR models significantly.
[ { "version": "v1", "created": "Mon, 27 Dec 2021 15:49:25 GMT" }, { "version": "v2", "created": "Wed, 16 Feb 2022 07:21:25 GMT" }, { "version": "v3", "created": "Tue, 16 Aug 2022 14:12:25 GMT" }, { "version": "v4", "created": "Wed, 17 Aug 2022 02:24:25 GMT" }, { "version": "v5", "created": "Tue, 1 Nov 2022 12:10:27 GMT" }, { "version": "v6", "created": "Thu, 30 Mar 2023 10:50:25 GMT" } ]
2023-03-31T00:00:00
[ [ "Pan", "Xiaofeng", "" ], [ "Shen", "Yibin", "" ], [ "Zhang", "Jing", "" ], [ "He", "Xu", "" ], [ "Huang", "Yang", "" ], [ "Wen", "Hong", "" ], [ "Mao", "Chengjun", "" ], [ "Cao", "Bo", "" ] ]
TITLE: MOEF: Modeling Occasion Evolution in Frequency Domain for Promotion-Aware Click-Through Rate Prediction ABSTRACT: Promotions are becoming more important and prevalent in e-commerce to attract customers and boost sales, leading to frequent changes of occasions, which drives users to behave differently. In such situations, most existing Click-Through Rate (CTR) models can't generalize well to online serving due to distribution uncertainty of the upcoming occasion. In this paper, we propose a novel CTR model named MOEF for recommendations under frequent changes of occasions. Firstly, we design a time series that consists of occasion signals generated from the online business scenario. Since occasion signals are more discriminative in the frequency domain, we apply Fourier Transformation to sliding time windows upon the time series, obtaining a sequence of frequency spectrum which is then processed by Occasion Evolution Layer (OEL). In this way, a high-order occasion representation can be learned to handle the online distribution uncertainty. Moreover, we adopt multiple experts to learn feature representations from multiple aspects, which are guided by the occasion representation via an attention mechanism. Accordingly, a mixture of feature representations is obtained adaptively for different occasions to predict the final CTR. Experimental results on real-world datasets validate the superiority of MOEF and online A/B tests also show MOEF outperforms representative CTR models significantly.
no_new_dataset
0.710025
2202.10650
Shixing Chen
Shixing Chen, Chun-Hao Liu, Xiang Hao, Xiaohan Nie, Maxim Arap, Raffay Hamid
Movies2Scenes: Using Movie Metadata to Learn Scene Representation
Accepted to CVPR 2023
null
null
null
cs.CV
http://creativecommons.org/licenses/by-nc-nd/4.0/
Understanding scenes in movies is crucial for a variety of applications such as video moderation, search, and recommendation. However, labeling individual scenes is a time-consuming process. In contrast, movie level metadata (e.g., genre, synopsis, etc.) regularly gets produced as part of the film production process, and is therefore significantly more commonly available. In this work, we propose a novel contrastive learning approach that uses movie metadata to learn a general-purpose scene representation. Specifically, we use movie metadata to define a measure of movie similarity, and use it during contrastive learning to limit our search for positive scene-pairs to only the movies that are considered similar to each other. Our learned scene representation consistently outperforms existing state-of-the-art methods on a diverse set of tasks evaluated using multiple benchmark datasets. Notably, our learned representation offers an average improvement of 7.9% on the seven classification tasks and 9.7% improvement on the two regression tasks in LVU dataset. Furthermore, using a newly collected movie dataset, we present comparative results of our scene representation on a set of video moderation tasks to demonstrate its generalizability on previously less explored tasks.
[ { "version": "v1", "created": "Tue, 22 Feb 2022 03:31:33 GMT" }, { "version": "v2", "created": "Sat, 12 Mar 2022 03:08:46 GMT" }, { "version": "v3", "created": "Thu, 30 Mar 2023 00:51:47 GMT" } ]
2023-03-31T00:00:00
[ [ "Chen", "Shixing", "" ], [ "Liu", "Chun-Hao", "" ], [ "Hao", "Xiang", "" ], [ "Nie", "Xiaohan", "" ], [ "Arap", "Maxim", "" ], [ "Hamid", "Raffay", "" ] ]
TITLE: Movies2Scenes: Using Movie Metadata to Learn Scene Representation ABSTRACT: Understanding scenes in movies is crucial for a variety of applications such as video moderation, search, and recommendation. However, labeling individual scenes is a time-consuming process. In contrast, movie level metadata (e.g., genre, synopsis, etc.) regularly gets produced as part of the film production process, and is therefore significantly more commonly available. In this work, we propose a novel contrastive learning approach that uses movie metadata to learn a general-purpose scene representation. Specifically, we use movie metadata to define a measure of movie similarity, and use it during contrastive learning to limit our search for positive scene-pairs to only the movies that are considered similar to each other. Our learned scene representation consistently outperforms existing state-of-the-art methods on a diverse set of tasks evaluated using multiple benchmark datasets. Notably, our learned representation offers an average improvement of 7.9% on the seven classification tasks and 9.7% improvement on the two regression tasks in LVU dataset. Furthermore, using a newly collected movie dataset, we present comparative results of our scene representation on a set of video moderation tasks to demonstrate its generalizability on previously less explored tasks.
new_dataset
0.71464
2204.13738
Jiang Liu
Jiang Liu, Srivathsa Pasumarthi, Ben Duffy, Enhao Gong, Keshav Datta, Greg Zaharchuk
One Model to Synthesize Them All: Multi-contrast Multi-scale Transformer for Missing Data Imputation
IEEE TMI accepted final version
null
10.1109/TMI.2023.3261707
null
eess.IV cs.CV
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
Multi-contrast magnetic resonance imaging (MRI) is widely used in clinical practice as each contrast provides complementary information. However, the availability of each imaging contrast may vary amongst patients, which poses challenges to radiologists and automated image analysis algorithms. A general approach for tackling this problem is missing data imputation, which aims to synthesize the missing contrasts from existing ones. While several convolutional neural networks (CNN) based algorithms have been proposed, they suffer from the fundamental limitations of CNN models, such as the requirement for fixed numbers of input and output channels, the inability to capture long-range dependencies, and the lack of interpretability. In this work, we formulate missing data imputation as a sequence-to-sequence learning problem and propose a multi-contrast multi-scale Transformer (MMT), which can take any subset of input contrasts and synthesize those that are missing. MMT consists of a multi-scale Transformer encoder that builds hierarchical representations of inputs combined with a multi-scale Transformer decoder that generates the outputs in a coarse-to-fine fashion. The proposed multi-contrast Swin Transformer blocks can efficiently capture intra- and inter-contrast dependencies for accurate image synthesis. Moreover, MMT is inherently interpretable as it allows us to understand the importance of each input contrast in different regions by analyzing the in-built attention maps of Transformer blocks in the decoder. Extensive experiments on two large-scale multi-contrast MRI datasets demonstrate that MMT outperforms the state-of-the-art methods quantitatively and qualitatively.
[ { "version": "v1", "created": "Thu, 28 Apr 2022 18:49:27 GMT" }, { "version": "v2", "created": "Wed, 22 Mar 2023 18:30:11 GMT" }, { "version": "v3", "created": "Wed, 29 Mar 2023 19:05:39 GMT" } ]
2023-03-31T00:00:00
[ [ "Liu", "Jiang", "" ], [ "Pasumarthi", "Srivathsa", "" ], [ "Duffy", "Ben", "" ], [ "Gong", "Enhao", "" ], [ "Datta", "Keshav", "" ], [ "Zaharchuk", "Greg", "" ] ]
TITLE: One Model to Synthesize Them All: Multi-contrast Multi-scale Transformer for Missing Data Imputation ABSTRACT: Multi-contrast magnetic resonance imaging (MRI) is widely used in clinical practice as each contrast provides complementary information. However, the availability of each imaging contrast may vary amongst patients, which poses challenges to radiologists and automated image analysis algorithms. A general approach for tackling this problem is missing data imputation, which aims to synthesize the missing contrasts from existing ones. While several convolutional neural networks (CNN) based algorithms have been proposed, they suffer from the fundamental limitations of CNN models, such as the requirement for fixed numbers of input and output channels, the inability to capture long-range dependencies, and the lack of interpretability. In this work, we formulate missing data imputation as a sequence-to-sequence learning problem and propose a multi-contrast multi-scale Transformer (MMT), which can take any subset of input contrasts and synthesize those that are missing. MMT consists of a multi-scale Transformer encoder that builds hierarchical representations of inputs combined with a multi-scale Transformer decoder that generates the outputs in a coarse-to-fine fashion. The proposed multi-contrast Swin Transformer blocks can efficiently capture intra- and inter-contrast dependencies for accurate image synthesis. Moreover, MMT is inherently interpretable as it allows us to understand the importance of each input contrast in different regions by analyzing the in-built attention maps of Transformer blocks in the decoder. Extensive experiments on two large-scale multi-contrast MRI datasets demonstrate that MMT outperforms the state-of-the-art methods quantitatively and qualitatively.
no_new_dataset
0.710258
2205.00459
Qingyan Meng
Qingyan Meng, Mingqing Xiao, Shen Yan, Yisen Wang, Zhouchen Lin, Zhi-Quan Luo
Training High-Performance Low-Latency Spiking Neural Networks by Differentiation on Spike Representation
Accepted by CVPR 2022
null
null
null
cs.NE cs.LG
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
Spiking Neural Network (SNN) is a promising energy-efficient AI model when implemented on neuromorphic hardware. However, it is a challenge to efficiently train SNNs due to their non-differentiability. Most existing methods either suffer from high latency (i.e., long simulation time steps), or cannot achieve as high performance as Artificial Neural Networks (ANNs). In this paper, we propose the Differentiation on Spike Representation (DSR) method, which could achieve high performance that is competitive to ANNs yet with low latency. First, we encode the spike trains into spike representation using (weighted) firing rate coding. Based on the spike representation, we systematically derive that the spiking dynamics with common neural models can be represented as some sub-differentiable mapping. With this viewpoint, our proposed DSR method trains SNNs through gradients of the mapping and avoids the common non-differentiability problem in SNN training. Then we analyze the error when representing the specific mapping with the forward computation of the SNN. To reduce such error, we propose to train the spike threshold in each layer, and to introduce a new hyperparameter for the neural models. With these components, the DSR method can achieve state-of-the-art SNN performance with low latency on both static and neuromorphic datasets, including CIFAR-10, CIFAR-100, ImageNet, and DVS-CIFAR10.
[ { "version": "v1", "created": "Sun, 1 May 2022 12:44:49 GMT" }, { "version": "v2", "created": "Thu, 30 Mar 2023 07:12:36 GMT" } ]
2023-03-31T00:00:00
[ [ "Meng", "Qingyan", "" ], [ "Xiao", "Mingqing", "" ], [ "Yan", "Shen", "" ], [ "Wang", "Yisen", "" ], [ "Lin", "Zhouchen", "" ], [ "Luo", "Zhi-Quan", "" ] ]
TITLE: Training High-Performance Low-Latency Spiking Neural Networks by Differentiation on Spike Representation ABSTRACT: Spiking Neural Network (SNN) is a promising energy-efficient AI model when implemented on neuromorphic hardware. However, it is a challenge to efficiently train SNNs due to their non-differentiability. Most existing methods either suffer from high latency (i.e., long simulation time steps), or cannot achieve as high performance as Artificial Neural Networks (ANNs). In this paper, we propose the Differentiation on Spike Representation (DSR) method, which could achieve high performance that is competitive to ANNs yet with low latency. First, we encode the spike trains into spike representation using (weighted) firing rate coding. Based on the spike representation, we systematically derive that the spiking dynamics with common neural models can be represented as some sub-differentiable mapping. With this viewpoint, our proposed DSR method trains SNNs through gradients of the mapping and avoids the common non-differentiability problem in SNN training. Then we analyze the error when representing the specific mapping with the forward computation of the SNN. To reduce such error, we propose to train the spike threshold in each layer, and to introduce a new hyperparameter for the neural models. With these components, the DSR method can achieve state-of-the-art SNN performance with low latency on both static and neuromorphic datasets, including CIFAR-10, CIFAR-100, ImageNet, and DVS-CIFAR10.
no_new_dataset
0.709466
2205.13115
Jaemin Cho
Jaemin Cho, Seunghyun Yoon, Ajinkya Kale, Franck Dernoncourt, Trung Bui, Mohit Bansal
Fine-grained Image Captioning with CLIP Reward
NAACL Findings 2022
null
null
null
cs.CL cs.AI cs.CV
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
Modern image captioning models are usually trained with text similarity objectives. However, since reference captions in public datasets often describe the most salient common objects, models trained with text similarity objectives tend to ignore specific and detailed aspects of an image that distinguish it from others. Toward more descriptive and distinctive caption generation, we propose using CLIP, a multimodal encoder trained on huge image-text pairs from web, to calculate multimodal similarity and use it as a reward function. We also propose a simple finetuning strategy of the CLIP text encoder to improve grammar that does not require extra text annotation. This completely eliminates the need for reference captions during the reward computation. To comprehensively evaluate descriptive captions, we introduce FineCapEval, a new dataset for caption evaluation with fine-grained criteria: overall, background, object, relations. In our experiments on text-to-image retrieval and FineCapEval, the proposed CLIP-guided model generates more distinctive captions than the CIDEr-optimized model. We also show that our unsupervised grammar finetuning of the CLIP text encoder alleviates the degeneration problem of the naive CLIP reward. Lastly, we show human analysis where the annotators strongly prefer the CLIP reward to the CIDEr and MLE objectives according to various criteria. Code and Data: https://github.com/j-min/CLIP-Caption-Reward
[ { "version": "v1", "created": "Thu, 26 May 2022 02:46:09 GMT" }, { "version": "v2", "created": "Wed, 29 Mar 2023 18:26:34 GMT" } ]
2023-03-31T00:00:00
[ [ "Cho", "Jaemin", "" ], [ "Yoon", "Seunghyun", "" ], [ "Kale", "Ajinkya", "" ], [ "Dernoncourt", "Franck", "" ], [ "Bui", "Trung", "" ], [ "Bansal", "Mohit", "" ] ]
TITLE: Fine-grained Image Captioning with CLIP Reward ABSTRACT: Modern image captioning models are usually trained with text similarity objectives. However, since reference captions in public datasets often describe the most salient common objects, models trained with text similarity objectives tend to ignore specific and detailed aspects of an image that distinguish it from others. Toward more descriptive and distinctive caption generation, we propose using CLIP, a multimodal encoder trained on huge image-text pairs from web, to calculate multimodal similarity and use it as a reward function. We also propose a simple finetuning strategy of the CLIP text encoder to improve grammar that does not require extra text annotation. This completely eliminates the need for reference captions during the reward computation. To comprehensively evaluate descriptive captions, we introduce FineCapEval, a new dataset for caption evaluation with fine-grained criteria: overall, background, object, relations. In our experiments on text-to-image retrieval and FineCapEval, the proposed CLIP-guided model generates more distinctive captions than the CIDEr-optimized model. We also show that our unsupervised grammar finetuning of the CLIP text encoder alleviates the degeneration problem of the naive CLIP reward. Lastly, we show human analysis where the annotators strongly prefer the CLIP reward to the CIDEr and MLE objectives according to various criteria. Code and Data: https://github.com/j-min/CLIP-Caption-Reward
new_dataset
0.719029
2206.03484
Lingchen Meng
Lingchen Meng, Xiyang Dai, Yinpeng Chen, Pengchuan Zhang, Dongdong Chen, Mengchen Liu, Jianfeng Wang, Zuxuan Wu, Lu Yuan, Yu-Gang Jiang
Detection Hub: Unifying Object Detection Datasets via Query Adaptation on Language Embedding
CVPR camera ready
null
null
null
cs.CV
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
Combining multiple datasets enables performance boost on many computer vision tasks. But similar trend has not been witnessed in object detection when combining multiple datasets due to two inconsistencies among detection datasets: taxonomy difference and domain gap. In this paper, we address these challenges by a new design (named Detection Hub) that is dataset-aware and category-aligned. It not only mitigates the dataset inconsistency but also provides coherent guidance for the detector to learn across multiple datasets. In particular, the dataset-aware design is achieved by learning a dataset embedding that is used to adapt object queries as well as convolutional kernels in detection heads. The categories across datasets are semantically aligned into a unified space by replacing one-hot category representations with word embedding and leveraging the semantic coherence of language embedding. Detection Hub fulfills the benefits of large data on object detection. Experiments demonstrate that joint training on multiple datasets achieves significant performance gains over training on each dataset alone. Detection Hub further achieves SoTA performance on UODB benchmark with wide variety of datasets.
[ { "version": "v1", "created": "Tue, 7 Jun 2022 17:59:44 GMT" }, { "version": "v2", "created": "Wed, 29 Mar 2023 18:00:08 GMT" } ]
2023-03-31T00:00:00
[ [ "Meng", "Lingchen", "" ], [ "Dai", "Xiyang", "" ], [ "Chen", "Yinpeng", "" ], [ "Zhang", "Pengchuan", "" ], [ "Chen", "Dongdong", "" ], [ "Liu", "Mengchen", "" ], [ "Wang", "Jianfeng", "" ], [ "Wu", "Zuxuan", "" ], [ "Yuan", "Lu", "" ], [ "Jiang", "Yu-Gang", "" ] ]
TITLE: Detection Hub: Unifying Object Detection Datasets via Query Adaptation on Language Embedding ABSTRACT: Combining multiple datasets enables performance boost on many computer vision tasks. But similar trend has not been witnessed in object detection when combining multiple datasets due to two inconsistencies among detection datasets: taxonomy difference and domain gap. In this paper, we address these challenges by a new design (named Detection Hub) that is dataset-aware and category-aligned. It not only mitigates the dataset inconsistency but also provides coherent guidance for the detector to learn across multiple datasets. In particular, the dataset-aware design is achieved by learning a dataset embedding that is used to adapt object queries as well as convolutional kernels in detection heads. The categories across datasets are semantically aligned into a unified space by replacing one-hot category representations with word embedding and leveraging the semantic coherence of language embedding. Detection Hub fulfills the benefits of large data on object detection. Experiments demonstrate that joint training on multiple datasets achieves significant performance gains over training on each dataset alone. Detection Hub further achieves SoTA performance on UODB benchmark with wide variety of datasets.
no_new_dataset
0.71081
2206.07796
MD. Mahim Anjum Haque
Md Mahim Anjum Haque and Wasi Uddin Ahmad and Ismini Lourentzou and Chris Brown
FixEval: Execution-based Evaluation of Program Fixes for Programming Problems
null
null
null
null
cs.SE cs.LG
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
The complexity of modern software has led to a drastic increase in the time and cost associated with detecting and rectifying software bugs. In response, researchers have explored various methods to automatically generate fixes for buggy code. However, due to the large combinatorial space of possible fixes for any given bug, few tools and datasets are available to evaluate model-generated fixes effectively. To address this issue, we introduce FixEval, a benchmark comprising of buggy code submissions to competitive programming problems and their corresponding fixes. FixEval offers an extensive collection of unit tests to evaluate the correctness of model-generated program fixes and assess further information regarding time, memory constraints, and acceptance based on a verdict. We consider two Transformer language models pretrained on programming languages as our baseline and compare them using match-based and execution-based evaluation metrics. Our experiments show that match-based metrics do not reflect model-generated program fixes accurately. At the same time, execution-based methods evaluate programs through all cases and scenarios designed explicitly for that solution. Therefore, we believe FixEval provides a step towards real-world automatic bug fixing and model-generated code evaluation. The dataset and models are open-sourced at https://github.com/mahimanzum/FixEval.
[ { "version": "v1", "created": "Wed, 15 Jun 2022 20:18:43 GMT" }, { "version": "v2", "created": "Fri, 26 Aug 2022 04:27:34 GMT" }, { "version": "v3", "created": "Thu, 29 Sep 2022 21:10:13 GMT" }, { "version": "v4", "created": "Thu, 30 Mar 2023 14:30:46 GMT" } ]
2023-03-31T00:00:00
[ [ "Haque", "Md Mahim Anjum", "" ], [ "Ahmad", "Wasi Uddin", "" ], [ "Lourentzou", "Ismini", "" ], [ "Brown", "Chris", "" ] ]
TITLE: FixEval: Execution-based Evaluation of Program Fixes for Programming Problems ABSTRACT: The complexity of modern software has led to a drastic increase in the time and cost associated with detecting and rectifying software bugs. In response, researchers have explored various methods to automatically generate fixes for buggy code. However, due to the large combinatorial space of possible fixes for any given bug, few tools and datasets are available to evaluate model-generated fixes effectively. To address this issue, we introduce FixEval, a benchmark comprising of buggy code submissions to competitive programming problems and their corresponding fixes. FixEval offers an extensive collection of unit tests to evaluate the correctness of model-generated program fixes and assess further information regarding time, memory constraints, and acceptance based on a verdict. We consider two Transformer language models pretrained on programming languages as our baseline and compare them using match-based and execution-based evaluation metrics. Our experiments show that match-based metrics do not reflect model-generated program fixes accurately. At the same time, execution-based methods evaluate programs through all cases and scenarios designed explicitly for that solution. Therefore, we believe FixEval provides a step towards real-world automatic bug fixing and model-generated code evaluation. The dataset and models are open-sourced at https://github.com/mahimanzum/FixEval.
new_dataset
0.712282
2206.08317
Zhifu Gao
Zhifu Gao, Shiliang Zhang, Ian McLoughlin, Zhijie Yan
Paraformer: Fast and Accurate Parallel Transformer for Non-autoregressive End-to-End Speech Recognition
5 pages, 3 figures, accepted by INTERSPEECH 2022
null
null
null
cs.SD cs.CL eess.AS
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
Transformers have recently dominated the ASR field. Although able to yield good performance, they involve an autoregressive (AR) decoder to generate tokens one by one, which is computationally inefficient. To speed up inference, non-autoregressive (NAR) methods, e.g. single-step NAR, were designed, to enable parallel generation. However, due to an independence assumption within the output tokens, performance of single-step NAR is inferior to that of AR models, especially with a large-scale corpus. There are two challenges to improving single-step NAR: Firstly to accurately predict the number of output tokens and extract hidden variables; secondly, to enhance modeling of interdependence between output tokens. To tackle both challenges, we propose a fast and accurate parallel transformer, termed Paraformer. This utilizes a continuous integrate-and-fire based predictor to predict the number of tokens and generate hidden variables. A glancing language model (GLM) sampler then generates semantic embeddings to enhance the NAR decoder's ability to model context interdependence. Finally, we design a strategy to generate negative samples for minimum word error rate training to further improve performance. Experiments using the public AISHELL-1, AISHELL-2 benchmark, and an industrial-level 20,000 hour task demonstrate that the proposed Paraformer can attain comparable performance to the state-of-the-art AR transformer, with more than 10x speedup.
[ { "version": "v1", "created": "Thu, 16 Jun 2022 17:24:14 GMT" }, { "version": "v2", "created": "Mon, 20 Jun 2022 09:34:48 GMT" }, { "version": "v3", "created": "Thu, 30 Mar 2023 07:00:38 GMT" } ]
2023-03-31T00:00:00
[ [ "Gao", "Zhifu", "" ], [ "Zhang", "Shiliang", "" ], [ "McLoughlin", "Ian", "" ], [ "Yan", "Zhijie", "" ] ]
TITLE: Paraformer: Fast and Accurate Parallel Transformer for Non-autoregressive End-to-End Speech Recognition ABSTRACT: Transformers have recently dominated the ASR field. Although able to yield good performance, they involve an autoregressive (AR) decoder to generate tokens one by one, which is computationally inefficient. To speed up inference, non-autoregressive (NAR) methods, e.g. single-step NAR, were designed, to enable parallel generation. However, due to an independence assumption within the output tokens, performance of single-step NAR is inferior to that of AR models, especially with a large-scale corpus. There are two challenges to improving single-step NAR: Firstly to accurately predict the number of output tokens and extract hidden variables; secondly, to enhance modeling of interdependence between output tokens. To tackle both challenges, we propose a fast and accurate parallel transformer, termed Paraformer. This utilizes a continuous integrate-and-fire based predictor to predict the number of tokens and generate hidden variables. A glancing language model (GLM) sampler then generates semantic embeddings to enhance the NAR decoder's ability to model context interdependence. Finally, we design a strategy to generate negative samples for minimum word error rate training to further improve performance. Experiments using the public AISHELL-1, AISHELL-2 benchmark, and an industrial-level 20,000 hour task demonstrate that the proposed Paraformer can attain comparable performance to the state-of-the-art AR transformer, with more than 10x speedup.
no_new_dataset
0.707847
2209.02970
Junjie Wang
Jiaxing Zhang, Ruyi Gan, Junjie Wang, Yuxiang Zhang, Lin Zhang, Ping Yang, Xinyu Gao, Ziwei Wu, Xiaoqun Dong, Junqing He, Jianheng Zhuo, Qi Yang, Yongfeng Huang, Xiayu Li, Yanghan Wu, Junyu Lu, Xinyu Zhu, Weifeng Chen, Ting Han, Kunhao Pan, Rui Wang, Hao Wang, Xiaojun Wu, Zhongshen Zeng, Chongpei Chen
Fengshenbang 1.0: Being the Foundation of Chinese Cognitive Intelligence
Added the Chinese version and is now a bilingual paper
null
null
null
cs.CL
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
Nowadays, foundation models become one of fundamental infrastructures in artificial intelligence, paving ways to the general intelligence. However, the reality presents two urgent challenges: existing foundation models are dominated by the English-language community; users are often given limited resources and thus cannot always use foundation models. To support the development of the Chinese-language community, we introduce an open-source project, called Fengshenbang, which leads by the research center for Cognitive Computing and Natural Language (CCNL). Our project has comprehensive capabilities, including large pre-trained models, user-friendly APIs, benchmarks, datasets, and others. We wrap all these in three sub-projects: the Fengshenbang Model, the Fengshen Framework, and the Fengshen Benchmark. An open-source roadmap, Fengshenbang, aims to re-evaluate the open-source community of Chinese pre-trained large-scale models, prompting the development of the entire Chinese large-scale model community. We also want to build a user-centered open-source ecosystem to allow individuals to access the desired models to match their computing resources. Furthermore, we invite companies, colleges, and research institutions to collaborate with us to build the large-scale open-source model-based ecosystem. We hope that this project will be the foundation of Chinese cognitive intelligence.
[ { "version": "v1", "created": "Wed, 7 Sep 2022 07:32:37 GMT" }, { "version": "v2", "created": "Thu, 6 Oct 2022 07:57:41 GMT" }, { "version": "v3", "created": "Thu, 30 Mar 2023 14:22:55 GMT" } ]
2023-03-31T00:00:00
[ [ "Zhang", "Jiaxing", "" ], [ "Gan", "Ruyi", "" ], [ "Wang", "Junjie", "" ], [ "Zhang", "Yuxiang", "" ], [ "Zhang", "Lin", "" ], [ "Yang", "Ping", "" ], [ "Gao", "Xinyu", "" ], [ "Wu", "Ziwei", "" ], [ "Dong", "Xiaoqun", "" ], [ "He", "Junqing", "" ], [ "Zhuo", "Jianheng", "" ], [ "Yang", "Qi", "" ], [ "Huang", "Yongfeng", "" ], [ "Li", "Xiayu", "" ], [ "Wu", "Yanghan", "" ], [ "Lu", "Junyu", "" ], [ "Zhu", "Xinyu", "" ], [ "Chen", "Weifeng", "" ], [ "Han", "Ting", "" ], [ "Pan", "Kunhao", "" ], [ "Wang", "Rui", "" ], [ "Wang", "Hao", "" ], [ "Wu", "Xiaojun", "" ], [ "Zeng", "Zhongshen", "" ], [ "Chen", "Chongpei", "" ] ]
TITLE: Fengshenbang 1.0: Being the Foundation of Chinese Cognitive Intelligence ABSTRACT: Nowadays, foundation models become one of fundamental infrastructures in artificial intelligence, paving ways to the general intelligence. However, the reality presents two urgent challenges: existing foundation models are dominated by the English-language community; users are often given limited resources and thus cannot always use foundation models. To support the development of the Chinese-language community, we introduce an open-source project, called Fengshenbang, which leads by the research center for Cognitive Computing and Natural Language (CCNL). Our project has comprehensive capabilities, including large pre-trained models, user-friendly APIs, benchmarks, datasets, and others. We wrap all these in three sub-projects: the Fengshenbang Model, the Fengshen Framework, and the Fengshen Benchmark. An open-source roadmap, Fengshenbang, aims to re-evaluate the open-source community of Chinese pre-trained large-scale models, prompting the development of the entire Chinese large-scale model community. We also want to build a user-centered open-source ecosystem to allow individuals to access the desired models to match their computing resources. Furthermore, we invite companies, colleges, and research institutions to collaborate with us to build the large-scale open-source model-based ecosystem. We hope that this project will be the foundation of Chinese cognitive intelligence.
no_new_dataset
0.705785
2209.04309
Jiawei Zheng
Jiawei Zheng and Petros Papapanagiotou and Jacques D. Fleuriot
Alignment-based conformance checking over probabilistic events
Extended version
null
null
null
cs.AI
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
Conformance checking techniques allow us to evaluate how well some exhibited behaviour, represented by a trace of monitored events, conforms to a specified process model. Modern monitoring and activity recognition technologies, such as those relying on sensors, the IoT, statistics and AI, can produce a wealth of relevant event data. However, this data is typically characterised by noise and uncertainty, in contrast to the assumption of a deterministic event log required by conformance checking algorithms. In this paper, we extend alignment-based conformance checking to function under a probabilistic event log. We introduce a weighted trace model and weighted alignment cost function, and a custom threshold parameter that controls the level of confidence on the event data vs. the process model. The resulting algorithm considers activities of lower but sufficiently high probability that better align with the process model. We explain the algorithm and its motivation both from formal and intuitive perspectives, and demonstrate its functionality in comparison with deterministic alignment using real-life datasets.
[ { "version": "v1", "created": "Fri, 9 Sep 2022 14:07:37 GMT" }, { "version": "v2", "created": "Thu, 30 Mar 2023 14:16:27 GMT" } ]
2023-03-31T00:00:00
[ [ "Zheng", "Jiawei", "" ], [ "Papapanagiotou", "Petros", "" ], [ "Fleuriot", "Jacques D.", "" ] ]
TITLE: Alignment-based conformance checking over probabilistic events ABSTRACT: Conformance checking techniques allow us to evaluate how well some exhibited behaviour, represented by a trace of monitored events, conforms to a specified process model. Modern monitoring and activity recognition technologies, such as those relying on sensors, the IoT, statistics and AI, can produce a wealth of relevant event data. However, this data is typically characterised by noise and uncertainty, in contrast to the assumption of a deterministic event log required by conformance checking algorithms. In this paper, we extend alignment-based conformance checking to function under a probabilistic event log. We introduce a weighted trace model and weighted alignment cost function, and a custom threshold parameter that controls the level of confidence on the event data vs. the process model. The resulting algorithm considers activities of lower but sufficiently high probability that better align with the process model. We explain the algorithm and its motivation both from formal and intuitive perspectives, and demonstrate its functionality in comparison with deterministic alignment using real-life datasets.
no_new_dataset
0.712182
2209.15200
Kai Li
Kai Li, Runxuan Yang, Xiaolin Hu
An efficient encoder-decoder architecture with top-down attention for speech separation
Accepted by ICLR 2023; Code & Demos: https://cslikai.cn/project/TDANet/
null
null
null
cs.SD cs.LG eess.AS
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
Deep neural networks have shown excellent prospects in speech separation tasks. However, obtaining good results while keeping a low model complexity remains challenging in real-world applications. In this paper, we provide a bio-inspired efficient encoder-decoder architecture by mimicking the brain's top-down attention, called TDANet, with decreased model complexity without sacrificing performance. The top-down attention in TDANet is extracted by the global attention (GA) module and the cascaded local attention (LA) layers. The GA module takes multi-scale acoustic features as input to extract global attention signal, which then modulates features of different scales by direct top-down connections. The LA layers use features of adjacent layers as input to extract the local attention signal, which is used to modulate the lateral input in a top-down manner. On three benchmark datasets, TDANet consistently achieved competitive separation performance to previous state-of-the-art (SOTA) methods with higher efficiency. Specifically, TDANet's multiply-accumulate operations (MACs) are only 5\% of Sepformer, one of the previous SOTA models, and CPU inference time is only 10\% of Sepformer. In addition, a large-size version of TDANet obtained SOTA results on three datasets, with MACs still only 10\% of Sepformer and the CPU inference time only 24\% of Sepformer.
[ { "version": "v1", "created": "Fri, 30 Sep 2022 03:09:53 GMT" }, { "version": "v2", "created": "Fri, 7 Oct 2022 06:42:03 GMT" }, { "version": "v3", "created": "Mon, 24 Oct 2022 14:52:49 GMT" }, { "version": "v4", "created": "Sat, 11 Feb 2023 17:44:23 GMT" }, { "version": "v5", "created": "Thu, 30 Mar 2023 06:01:28 GMT" } ]
2023-03-31T00:00:00
[ [ "Li", "Kai", "" ], [ "Yang", "Runxuan", "" ], [ "Hu", "Xiaolin", "" ] ]
TITLE: An efficient encoder-decoder architecture with top-down attention for speech separation ABSTRACT: Deep neural networks have shown excellent prospects in speech separation tasks. However, obtaining good results while keeping a low model complexity remains challenging in real-world applications. In this paper, we provide a bio-inspired efficient encoder-decoder architecture by mimicking the brain's top-down attention, called TDANet, with decreased model complexity without sacrificing performance. The top-down attention in TDANet is extracted by the global attention (GA) module and the cascaded local attention (LA) layers. The GA module takes multi-scale acoustic features as input to extract global attention signal, which then modulates features of different scales by direct top-down connections. The LA layers use features of adjacent layers as input to extract the local attention signal, which is used to modulate the lateral input in a top-down manner. On three benchmark datasets, TDANet consistently achieved competitive separation performance to previous state-of-the-art (SOTA) methods with higher efficiency. Specifically, TDANet's multiply-accumulate operations (MACs) are only 5\% of Sepformer, one of the previous SOTA models, and CPU inference time is only 10\% of Sepformer. In addition, a large-size version of TDANet obtained SOTA results on three datasets, with MACs still only 10\% of Sepformer and the CPU inference time only 24\% of Sepformer.
no_new_dataset
0.709453
2210.06756
Huiguang He
Changde Du, Kaicheng Fu, Jinpeng Li, Huiguang He
Decoding Visual Neural Representations by Multimodal Learning of Brain-Visual-Linguistic Features
IEEE Transactions on Pattern Analysis and Machine Intelligence (TPAMI)
null
null
null
cs.CV cs.AI cs.MM cs.NE
http://creativecommons.org/licenses/by/4.0/
Decoding human visual neural representations is a challenging task with great scientific significance in revealing vision-processing mechanisms and developing brain-like intelligent machines. Most existing methods are difficult to generalize to novel categories that have no corresponding neural data for training. The two main reasons are 1) the under-exploitation of the multimodal semantic knowledge underlying the neural data and 2) the small number of paired (stimuli-responses) training data. To overcome these limitations, this paper presents a generic neural decoding method called BraVL that uses multimodal learning of brain-visual-linguistic features. We focus on modeling the relationships between brain, visual and linguistic features via multimodal deep generative models. Specifically, we leverage the mixture-of-product-of-experts formulation to infer a latent code that enables a coherent joint generation of all three modalities. To learn a more consistent joint representation and improve the data efficiency in the case of limited brain activity data, we exploit both intra- and inter-modality mutual information maximization regularization terms. In particular, our BraVL model can be trained under various semi-supervised scenarios to incorporate the visual and textual features obtained from the extra categories. Finally, we construct three trimodal matching datasets, and the extensive experiments lead to some interesting conclusions and cognitive insights: 1) decoding novel visual categories from human brain activity is practically possible with good accuracy; 2) decoding models using the combination of visual and linguistic features perform much better than those using either of them alone; 3) visual perception may be accompanied by linguistic influences to represent the semantics of visual stimuli. Code and data: https://github.com/ChangdeDu/BraVL.
[ { "version": "v1", "created": "Thu, 13 Oct 2022 05:49:33 GMT" }, { "version": "v2", "created": "Thu, 30 Mar 2023 15:27:33 GMT" } ]
2023-03-31T00:00:00
[ [ "Du", "Changde", "" ], [ "Fu", "Kaicheng", "" ], [ "Li", "Jinpeng", "" ], [ "He", "Huiguang", "" ] ]
TITLE: Decoding Visual Neural Representations by Multimodal Learning of Brain-Visual-Linguistic Features ABSTRACT: Decoding human visual neural representations is a challenging task with great scientific significance in revealing vision-processing mechanisms and developing brain-like intelligent machines. Most existing methods are difficult to generalize to novel categories that have no corresponding neural data for training. The two main reasons are 1) the under-exploitation of the multimodal semantic knowledge underlying the neural data and 2) the small number of paired (stimuli-responses) training data. To overcome these limitations, this paper presents a generic neural decoding method called BraVL that uses multimodal learning of brain-visual-linguistic features. We focus on modeling the relationships between brain, visual and linguistic features via multimodal deep generative models. Specifically, we leverage the mixture-of-product-of-experts formulation to infer a latent code that enables a coherent joint generation of all three modalities. To learn a more consistent joint representation and improve the data efficiency in the case of limited brain activity data, we exploit both intra- and inter-modality mutual information maximization regularization terms. In particular, our BraVL model can be trained under various semi-supervised scenarios to incorporate the visual and textual features obtained from the extra categories. Finally, we construct three trimodal matching datasets, and the extensive experiments lead to some interesting conclusions and cognitive insights: 1) decoding novel visual categories from human brain activity is practically possible with good accuracy; 2) decoding models using the combination of visual and linguistic features perform much better than those using either of them alone; 3) visual perception may be accompanied by linguistic influences to represent the semantics of visual stimuli. Code and data: https://github.com/ChangdeDu/BraVL.
no_new_dataset
0.709585
2210.11588
Desh Raj
Desh Raj, Junteng Jia, Jay Mahadeokar, Chunyang Wu, Niko Moritz, Xiaohui Zhang, Ozlem Kalinli
Anchored Speech Recognition with Neural Transducers
To appear at IEEE ICASSP 2023
null
null
null
eess.AS cs.SD
http://creativecommons.org/licenses/by/4.0/
Neural transducers have achieved human level performance on standard speech recognition benchmarks. However, their performance significantly degrades in the presence of cross-talk, especially when the primary speaker has a low signal-to-noise ratio. Anchored speech recognition refers to a class of methods that use information from an anchor segment (e.g., wake-words) to recognize device-directed speech while ignoring interfering background speech. In this paper, we investigate anchored speech recognition to make neural transducers robust to background speech. We extract context information from the anchor segment with a tiny auxiliary network, and use encoder biasing and joiner gating to guide the transducer towards the target speech. Moreover, to improve the robustness of context embedding extraction, we propose auxiliary training objectives to disentangle lexical content from speaking style. We evaluate our methods on synthetic LibriSpeech-based mixtures comprising several SNR and overlap conditions; they improve relative word error rates by 19.6% over a strong baseline, when averaged over all conditions.
[ { "version": "v1", "created": "Thu, 20 Oct 2022 21:00:42 GMT" }, { "version": "v2", "created": "Sun, 12 Mar 2023 18:33:03 GMT" }, { "version": "v3", "created": "Wed, 29 Mar 2023 18:17:13 GMT" } ]
2023-03-31T00:00:00
[ [ "Raj", "Desh", "" ], [ "Jia", "Junteng", "" ], [ "Mahadeokar", "Jay", "" ], [ "Wu", "Chunyang", "" ], [ "Moritz", "Niko", "" ], [ "Zhang", "Xiaohui", "" ], [ "Kalinli", "Ozlem", "" ] ]
TITLE: Anchored Speech Recognition with Neural Transducers ABSTRACT: Neural transducers have achieved human level performance on standard speech recognition benchmarks. However, their performance significantly degrades in the presence of cross-talk, especially when the primary speaker has a low signal-to-noise ratio. Anchored speech recognition refers to a class of methods that use information from an anchor segment (e.g., wake-words) to recognize device-directed speech while ignoring interfering background speech. In this paper, we investigate anchored speech recognition to make neural transducers robust to background speech. We extract context information from the anchor segment with a tiny auxiliary network, and use encoder biasing and joiner gating to guide the transducer towards the target speech. Moreover, to improve the robustness of context embedding extraction, we propose auxiliary training objectives to disentangle lexical content from speaking style. We evaluate our methods on synthetic LibriSpeech-based mixtures comprising several SNR and overlap conditions; they improve relative word error rates by 19.6% over a strong baseline, when averaged over all conditions.
no_new_dataset
0.710415
2210.11624
Houssem Meghnoudj
Houssem Meghnoudj (1), Bogdan Robu (1), Mazen Alamir (1) ((1) Univ. Grenoble Alpes, CNRS, Grenoble INP, GIPSA-lab, 38000 Grenoble, France)
Sparse Dynamical Features generation, application to Parkinson's Disease diagnosis
18 pages, 13 figures
null
null
null
eess.SY cs.LG cs.SY
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
In this study we focus on the diagnosis of Parkinson's Disease (PD) based on electroencephalogram (EEG) signals. We propose a new approach inspired by the functioning of the brain that uses the dynamics, frequency and temporal content of EEGs to extract new demarcating features of the disease. The method was evaluated on a publicly available dataset containing EEG signals recorded during a 3-oddball auditory task involving N = 50 subjects, of whom 25 suffer from PD. By extracting two features, and separating them with a straight line using a Linear Discriminant Analysis (LDA) classifier, we can separate the healthy from the unhealthy subjects with an accuracy of 90 % $(p < 0.03)$ using a single channel. By aggregating the information from three channels and making them vote, we obtain an accuracy of 94 %, a sensitivity of 96 % and a specificity of 92 %. The evaluation was carried out using a nested Leave-One-Out cross-validation procedure, thus preventing data leakage problems and giving a less biased evaluation. Several tests were carried out to assess the validity and robustness of our approach, including the test where we use only half the available data for training. Under this constraint, the model achieves an accuracy of 83.8 %.
[ { "version": "v1", "created": "Thu, 20 Oct 2022 22:39:29 GMT" }, { "version": "v2", "created": "Wed, 29 Mar 2023 15:04:50 GMT" } ]
2023-03-31T00:00:00
[ [ "Meghnoudj", "Houssem", "" ], [ "Robu", "Bogdan", "" ], [ "Alamir", "Mazen", "" ] ]
TITLE: Sparse Dynamical Features generation, application to Parkinson's Disease diagnosis ABSTRACT: In this study we focus on the diagnosis of Parkinson's Disease (PD) based on electroencephalogram (EEG) signals. We propose a new approach inspired by the functioning of the brain that uses the dynamics, frequency and temporal content of EEGs to extract new demarcating features of the disease. The method was evaluated on a publicly available dataset containing EEG signals recorded during a 3-oddball auditory task involving N = 50 subjects, of whom 25 suffer from PD. By extracting two features, and separating them with a straight line using a Linear Discriminant Analysis (LDA) classifier, we can separate the healthy from the unhealthy subjects with an accuracy of 90 % $(p < 0.03)$ using a single channel. By aggregating the information from three channels and making them vote, we obtain an accuracy of 94 %, a sensitivity of 96 % and a specificity of 92 %. The evaluation was carried out using a nested Leave-One-Out cross-validation procedure, thus preventing data leakage problems and giving a less biased evaluation. Several tests were carried out to assess the validity and robustness of our approach, including the test where we use only half the available data for training. Under this constraint, the model achieves an accuracy of 83.8 %.
no_new_dataset
0.707859
2210.15430
Varun Mandalapu
Varun Mandalapu, Lujie Karen Chen, Sushruta Shetty, Zhiyuan Chen, Jiaqi Gong
Student-centric Model of Learning Management System Activity and Academic Performance: from Correlation to Causation
null
null
null
null
cs.CY cs.AI cs.LG
http://creativecommons.org/licenses/by-nc-nd/4.0/
In recent years, there is a lot of interest in modeling students' digital traces in Learning Management System (LMS) to understand students' learning behavior patterns including aspects of meta-cognition and self-regulation, with the ultimate goal to turn those insights into actionable information to support students to improve their learning outcomes. In achieving this goal, however, there are two main issues that need to be addressed given the existing literature. Firstly, most of the current work is course-centered (i.e. models are built from data for a specific course) rather than student-centered; secondly, a vast majority of the models are correlational rather than causal. Those issues make it challenging to identify the most promising actionable factors for intervention at the student level where most of the campus-wide academic support is designed for. In this paper, we explored a student-centric analytical framework for LMS activity data that can provide not only correlational but causal insights mined from observational data. We demonstrated this approach using a dataset of 1651 computing major students at a public university in the US during one semester in the Fall of 2019. This dataset includes students' fine-grained LMS interaction logs and administrative data, e.g. demographics and academic performance. In addition, we expand the repository of LMS behavior indicators to include those that can characterize the time-of-the-day of login (e.g. chronotype). Our analysis showed that student login volume, compared with other login behavior indicators, is both strongly correlated and causally linked to student academic performance, especially among students with low academic performance. We envision that those insights will provide convincing evidence for college student support groups to launch student-centered and targeted interventions that are effective and scalable.
[ { "version": "v1", "created": "Thu, 27 Oct 2022 14:08:25 GMT" }, { "version": "v2", "created": "Tue, 31 Jan 2023 23:59:22 GMT" }, { "version": "v3", "created": "Thu, 30 Mar 2023 01:08:17 GMT" } ]
2023-03-31T00:00:00
[ [ "Mandalapu", "Varun", "" ], [ "Chen", "Lujie Karen", "" ], [ "Shetty", "Sushruta", "" ], [ "Chen", "Zhiyuan", "" ], [ "Gong", "Jiaqi", "" ] ]
TITLE: Student-centric Model of Learning Management System Activity and Academic Performance: from Correlation to Causation ABSTRACT: In recent years, there is a lot of interest in modeling students' digital traces in Learning Management System (LMS) to understand students' learning behavior patterns including aspects of meta-cognition and self-regulation, with the ultimate goal to turn those insights into actionable information to support students to improve their learning outcomes. In achieving this goal, however, there are two main issues that need to be addressed given the existing literature. Firstly, most of the current work is course-centered (i.e. models are built from data for a specific course) rather than student-centered; secondly, a vast majority of the models are correlational rather than causal. Those issues make it challenging to identify the most promising actionable factors for intervention at the student level where most of the campus-wide academic support is designed for. In this paper, we explored a student-centric analytical framework for LMS activity data that can provide not only correlational but causal insights mined from observational data. We demonstrated this approach using a dataset of 1651 computing major students at a public university in the US during one semester in the Fall of 2019. This dataset includes students' fine-grained LMS interaction logs and administrative data, e.g. demographics and academic performance. In addition, we expand the repository of LMS behavior indicators to include those that can characterize the time-of-the-day of login (e.g. chronotype). Our analysis showed that student login volume, compared with other login behavior indicators, is both strongly correlated and causally linked to student academic performance, especially among students with low academic performance. We envision that those insights will provide convincing evidence for college student support groups to launch student-centered and targeted interventions that are effective and scalable.
no_new_dataset
0.696191
2210.15511
Zhi-Qi Cheng
Jin-Peng Lan, Zhi-Qi Cheng, Jun-Yan He, Chenyang Li, Bin Luo, Xu Bao, Wangmeng Xiang, Yifeng Geng, Xuansong Xie
ProContEXT: Exploring Progressive Context Transformer for Tracking
Accepted at ICASSP 2023, source code is at https://github.com/zhiqic/ProContEXT
null
null
null
cs.CV cs.AI cs.MM
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
Existing Visual Object Tracking (VOT) only takes the target area in the first frame as a template. This causes tracking to inevitably fail in fast-changing and crowded scenes, as it cannot account for changes in object appearance between frames. To this end, we revamped the tracking framework with Progressive Context Encoding Transformer Tracker (ProContEXT), which coherently exploits spatial and temporal contexts to predict object motion trajectories. Specifically, ProContEXT leverages a context-aware self-attention module to encode the spatial and temporal context, refining and updating the multi-scale static and dynamic templates to progressively perform accurately tracking. It explores the complementary between spatial and temporal context, raising a new pathway to multi-context modeling for transformer-based trackers. In addition, ProContEXT revised the token pruning technique to reduce computational complexity. Extensive experiments on popular benchmark datasets such as GOT-10k and TrackingNet demonstrate that the proposed ProContEXT achieves state-of-the-art performance.
[ { "version": "v1", "created": "Thu, 27 Oct 2022 14:47:19 GMT" }, { "version": "v2", "created": "Wed, 23 Nov 2022 21:45:32 GMT" }, { "version": "v3", "created": "Mon, 27 Mar 2023 02:02:43 GMT" }, { "version": "v4", "created": "Thu, 30 Mar 2023 06:12:26 GMT" } ]
2023-03-31T00:00:00
[ [ "Lan", "Jin-Peng", "" ], [ "Cheng", "Zhi-Qi", "" ], [ "He", "Jun-Yan", "" ], [ "Li", "Chenyang", "" ], [ "Luo", "Bin", "" ], [ "Bao", "Xu", "" ], [ "Xiang", "Wangmeng", "" ], [ "Geng", "Yifeng", "" ], [ "Xie", "Xuansong", "" ] ]
TITLE: ProContEXT: Exploring Progressive Context Transformer for Tracking ABSTRACT: Existing Visual Object Tracking (VOT) only takes the target area in the first frame as a template. This causes tracking to inevitably fail in fast-changing and crowded scenes, as it cannot account for changes in object appearance between frames. To this end, we revamped the tracking framework with Progressive Context Encoding Transformer Tracker (ProContEXT), which coherently exploits spatial and temporal contexts to predict object motion trajectories. Specifically, ProContEXT leverages a context-aware self-attention module to encode the spatial and temporal context, refining and updating the multi-scale static and dynamic templates to progressively perform accurately tracking. It explores the complementary between spatial and temporal context, raising a new pathway to multi-context modeling for transformer-based trackers. In addition, ProContEXT revised the token pruning technique to reduce computational complexity. Extensive experiments on popular benchmark datasets such as GOT-10k and TrackingNet demonstrate that the proposed ProContEXT achieves state-of-the-art performance.
no_new_dataset
0.710258
2210.15518
Zhi-Qi Cheng
Chenyang Li, Zhi-Qi Cheng, Jun-Yan He, Pengyu Li, Bin Luo, Hanyuan Chen, Yifeng Geng, Jin-Peng Lan, Xuansong Xie
LongShortNet: Exploring Temporal and Semantic Features Fusion in Streaming Perception
Accepted at ICASSP 2023, source code is at https://github.com/zhiqic/LongShortNet
null
null
null
cs.CV cs.AI cs.MM
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
Streaming perception is a critical task in autonomous driving that requires balancing the latency and accuracy of the autopilot system. However, current methods for streaming perception are limited as they only rely on the current and adjacent two frames to learn movement patterns. This restricts their ability to model complex scenes, often resulting in poor detection results. To address this limitation, we propose LongShortNet, a novel dual-path network that captures long-term temporal motion and integrates it with short-term spatial semantics for real-time perception. LongShortNet is notable as it is the first work to extend long-term temporal modeling to streaming perception, enabling spatiotemporal feature fusion. We evaluate LongShortNet on the challenging Argoverse-HD dataset and demonstrate that it outperforms existing state-of-the-art methods with almost no additional computational cost.
[ { "version": "v1", "created": "Thu, 27 Oct 2022 14:57:14 GMT" }, { "version": "v2", "created": "Wed, 23 Nov 2022 13:26:22 GMT" }, { "version": "v3", "created": "Mon, 27 Mar 2023 02:08:57 GMT" }, { "version": "v4", "created": "Thu, 30 Mar 2023 04:02:18 GMT" } ]
2023-03-31T00:00:00
[ [ "Li", "Chenyang", "" ], [ "Cheng", "Zhi-Qi", "" ], [ "He", "Jun-Yan", "" ], [ "Li", "Pengyu", "" ], [ "Luo", "Bin", "" ], [ "Chen", "Hanyuan", "" ], [ "Geng", "Yifeng", "" ], [ "Lan", "Jin-Peng", "" ], [ "Xie", "Xuansong", "" ] ]
TITLE: LongShortNet: Exploring Temporal and Semantic Features Fusion in Streaming Perception ABSTRACT: Streaming perception is a critical task in autonomous driving that requires balancing the latency and accuracy of the autopilot system. However, current methods for streaming perception are limited as they only rely on the current and adjacent two frames to learn movement patterns. This restricts their ability to model complex scenes, often resulting in poor detection results. To address this limitation, we propose LongShortNet, a novel dual-path network that captures long-term temporal motion and integrates it with short-term spatial semantics for real-time perception. LongShortNet is notable as it is the first work to extend long-term temporal modeling to streaming perception, enabling spatiotemporal feature fusion. We evaluate LongShortNet on the challenging Argoverse-HD dataset and demonstrate that it outperforms existing state-of-the-art methods with almost no additional computational cost.
no_new_dataset
0.711193