Pub Date : 2021-10-01DOI: 10.1109/ICCV48922.2021.00881
Z. Gao, Shufei Zhang, Kaizhu Huang, Qiufeng Wang, Chaoliang Zhong
The latest heuristic for handling the domain shift in un-supervised domain adaptation tasks is to reduce the data distribution discrepancy using adversarial learning. Recent studies improve the conventional adversarial domain adaptation methods with discriminative information by integrating the classifier’s outputs into distribution divergence measurement. However, they still suffer from the equilibrium problem of adversarial learning in which even if the discriminator is fully confused, sufficient similarity between two distributions cannot be guaranteed. To overcome this problem, we propose a novel approach named feature gradient distribution alignment (FGDA)1. We demonstrate the rationale of our method both theoretically and empirically. In particular, we show that the distribution discrepancy can be reduced by constraining feature gradients of two domains to have similar distributions. Meanwhile, our method enjoys a theoretical guarantee that a tighter error upper bound for target samples can be obtained than that of conventional adversarial domain adaptation methods. By integrating the proposed method with existing adversarial domain adaptation models, we achieve state-of-the-art performance on two real-world benchmark datasets.
{"title":"Gradient Distribution Alignment Certificates Better Adversarial Domain Adaptation","authors":"Z. Gao, Shufei Zhang, Kaizhu Huang, Qiufeng Wang, Chaoliang Zhong","doi":"10.1109/ICCV48922.2021.00881","DOIUrl":"https://doi.org/10.1109/ICCV48922.2021.00881","url":null,"abstract":"The latest heuristic for handling the domain shift in un-supervised domain adaptation tasks is to reduce the data distribution discrepancy using adversarial learning. Recent studies improve the conventional adversarial domain adaptation methods with discriminative information by integrating the classifier’s outputs into distribution divergence measurement. However, they still suffer from the equilibrium problem of adversarial learning in which even if the discriminator is fully confused, sufficient similarity between two distributions cannot be guaranteed. To overcome this problem, we propose a novel approach named feature gradient distribution alignment (FGDA)1. We demonstrate the rationale of our method both theoretically and empirically. In particular, we show that the distribution discrepancy can be reduced by constraining feature gradients of two domains to have similar distributions. Meanwhile, our method enjoys a theoretical guarantee that a tighter error upper bound for target samples can be obtained than that of conventional adversarial domain adaptation methods. By integrating the proposed method with existing adversarial domain adaptation models, we achieve state-of-the-art performance on two real-world benchmark datasets.","PeriodicalId":6820,"journal":{"name":"2021 IEEE/CVF International Conference on Computer Vision (ICCV)","volume":"5 1","pages":"8917-8926"},"PeriodicalIF":0.0,"publicationDate":"2021-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"79687852","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Humans are able to continuously detect and track surrounding objects by constructing a spatial-temporal memory of the objects when looking around. In contrast, 3D object detectors in existing tracking-by-detection systems often search for objects in every new video frame from scratch, without fully leveraging memory from previous detection results. In this work, we propose a novel system for integrated 3D object detection and tracking, which uses a dynamic object occupancy map and previous object states as spatial-temporal memory to assist object detection in future frames. This memory, together with the ego-motion from back-end odometry, guides the detector to achieve more efficient object proposal generation and more accurate object state estimation. The experiments demonstrate the effectiveness of the proposed system and its performance on the ScanNet and KITTI datasets. Moreover, the proposed system produces stable bounding boxes and pose trajectories over time, while being able to handle occluded and truncated objects. Code is available at the project page: https://zju3dv.github.io/UDOLO.
{"title":"You Don’t Only Look Once: Constructing Spatial-Temporal Memory for Integrated 3D Object Detection and Tracking","authors":"Jiaming Sun, Yiming Xie, Siyu Zhang, Linghao Chen, Guofeng Zhang, H. Bao, Xiaowei Zhou","doi":"10.1109/ICCV48922.2021.00317","DOIUrl":"https://doi.org/10.1109/ICCV48922.2021.00317","url":null,"abstract":"Humans are able to continuously detect and track surrounding objects by constructing a spatial-temporal memory of the objects when looking around. In contrast, 3D object detectors in existing tracking-by-detection systems often search for objects in every new video frame from scratch, without fully leveraging memory from previous detection results. In this work, we propose a novel system for integrated 3D object detection and tracking, which uses a dynamic object occupancy map and previous object states as spatial-temporal memory to assist object detection in future frames. This memory, together with the ego-motion from back-end odometry, guides the detector to achieve more efficient object proposal generation and more accurate object state estimation. The experiments demonstrate the effectiveness of the proposed system and its performance on the ScanNet and KITTI datasets. Moreover, the proposed system produces stable bounding boxes and pose trajectories over time, while being able to handle occluded and truncated objects. Code is available at the project page: https://zju3dv.github.io/UDOLO.","PeriodicalId":6820,"journal":{"name":"2021 IEEE/CVF International Conference on Computer Vision (ICCV)","volume":"9 1","pages":"3265-3174"},"PeriodicalIF":0.0,"publicationDate":"2021-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"79829259","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-10-01DOI: 10.1109/ICCV48922.2021.00219
Jiahe Shi, Yali Li, Shengjin Wang
Human-oriented image captioning with both high diversity and accuracy is a challenging task in vision+language modeling. The reinforcement learning (RL) based frameworks promote the accuracy of image captioning, yet seriously hurt the diversity. In contrast, other methods based on variational auto-encoder (VAE) or generative adversarial network (GAN) can produce diverse yet less accurate captions. In this work, we devote our attention to promote the diversity of RL-based image captioning. To be specific, we devise a partial off-policy learning scheme to balance accuracy and diversity. First, we keep the model exposed to varied candidate captions by sampling from the initial state before RL launched. Second, a novel criterion named max-CIDEr is proposed to serve as the reward for promoting diversity. We combine the above-mentioned offpolicy strategy with the on-policy one to moderate the exploration effect, further balancing the diversity and accuracy for human-like image captioning. Experiments show that our method locates the closest to human performance in the diversity-accuracy space, and achieves the highest Pearson correlation as 0.337 with human performance.
{"title":"Partial Off-policy Learning: Balance Accuracy and Diversity for Human-Oriented Image Captioning","authors":"Jiahe Shi, Yali Li, Shengjin Wang","doi":"10.1109/ICCV48922.2021.00219","DOIUrl":"https://doi.org/10.1109/ICCV48922.2021.00219","url":null,"abstract":"Human-oriented image captioning with both high diversity and accuracy is a challenging task in vision+language modeling. The reinforcement learning (RL) based frameworks promote the accuracy of image captioning, yet seriously hurt the diversity. In contrast, other methods based on variational auto-encoder (VAE) or generative adversarial network (GAN) can produce diverse yet less accurate captions. In this work, we devote our attention to promote the diversity of RL-based image captioning. To be specific, we devise a partial off-policy learning scheme to balance accuracy and diversity. First, we keep the model exposed to varied candidate captions by sampling from the initial state before RL launched. Second, a novel criterion named max-CIDEr is proposed to serve as the reward for promoting diversity. We combine the above-mentioned offpolicy strategy with the on-policy one to moderate the exploration effect, further balancing the diversity and accuracy for human-like image captioning. Experiments show that our method locates the closest to human performance in the diversity-accuracy space, and achieves the highest Pearson correlation as 0.337 with human performance.","PeriodicalId":6820,"journal":{"name":"2021 IEEE/CVF International Conference on Computer Vision (ICCV)","volume":"42 1","pages":"2167-2176"},"PeriodicalIF":0.0,"publicationDate":"2021-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"82518379","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-10-01DOI: 10.1109/ICCV48922.2021.01238
Rasha Friji, Hassen Drira, F. Chaieb, Hamza Kchok, S. Kurtek
Deep Learning architectures, albeit successful in most computer vision tasks, were designed for data with an underlying Euclidean structure, which is not usually fulfilled since pre-processed data may lie on a non-linear space. In this paper, we propose a geometry aware deep learning approach using rigid and non rigid transformation optimization for skeleton-based action recognition. Skeleton sequences are first modeled as trajectories on Kendall’s shape space and then mapped to the linear tangent space. The resulting structured data are then fed to a deep learning architecture, which includes a layer that optimizes over rigid and non rigid transformations of the 3D skeletons, followed by a CNN-LSTM network. The assessment on two large scale skeleton datasets, namely NTU-RGB+D and NTU-RGB+D 120, has proven that the proposed approach outperforms existing geometric deep learning methods and exceeds recently published approaches with respect to the majority of configurations.
{"title":"Geometric Deep Neural Network using Rigid and Non-Rigid Transformations for Human Action Recognition","authors":"Rasha Friji, Hassen Drira, F. Chaieb, Hamza Kchok, S. Kurtek","doi":"10.1109/ICCV48922.2021.01238","DOIUrl":"https://doi.org/10.1109/ICCV48922.2021.01238","url":null,"abstract":"Deep Learning architectures, albeit successful in most computer vision tasks, were designed for data with an underlying Euclidean structure, which is not usually fulfilled since pre-processed data may lie on a non-linear space. In this paper, we propose a geometry aware deep learning approach using rigid and non rigid transformation optimization for skeleton-based action recognition. Skeleton sequences are first modeled as trajectories on Kendall’s shape space and then mapped to the linear tangent space. The resulting structured data are then fed to a deep learning architecture, which includes a layer that optimizes over rigid and non rigid transformations of the 3D skeletons, followed by a CNN-LSTM network. The assessment on two large scale skeleton datasets, namely NTU-RGB+D and NTU-RGB+D 120, has proven that the proposed approach outperforms existing geometric deep learning methods and exceeds recently published approaches with respect to the majority of configurations.","PeriodicalId":6820,"journal":{"name":"2021 IEEE/CVF International Conference on Computer Vision (ICCV)","volume":"20 1","pages":"12591-12600"},"PeriodicalIF":0.0,"publicationDate":"2021-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"80321482","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-10-01DOI: 10.1109/ICCV48922.2021.01503
Piaopiao Yu, Jie Guo, Fan Huang, Cheng Zhou, H. Che, Xiao Ling, Yanwen Guo
Data-driven sky models have gained much attention in outdoor illumination prediction recently, showing superior performance against analytical models. However, naively compressing an outdoor panorama into a low-dimensional latent vector, as existing models have done, causes two major problems. One is the mutual interference between the HDR intensity of the sun and the complex textures of the surrounding sky, and the other is the lack of fine-grained control over independent lighting factors due to the entangled representation. To address these issues, we propose a hierarchical disentangled sky model (HDSky) for outdoor illumination prediction. With this model, any outdoor panorama can be hierarchically disentangled into several factors based on three well-designed autoencoders. The first autoencoder compresses each sunny panorama into a sky vector and a sun vector with some constraints. The second autoencoder and the third autoencoder further disentangle the sun intensity and the sky intensity from the sun vector and the sky vector with several customized loss functions respectively. Moreover, a unified framework is designed to predict all-weather sky information from a single outdoor image. Through extensive experiments, we demonstrate that the proposed model significantly improves the accuracy of outdoor illumination prediction. It also allows users to intuitively edit the predicted panorama (e.g., changing the position of the sun while preserving others), without sacrificing physical plausibility.
{"title":"Hierarchical Disentangled Representation Learning for Outdoor Illumination Estimation and Editing","authors":"Piaopiao Yu, Jie Guo, Fan Huang, Cheng Zhou, H. Che, Xiao Ling, Yanwen Guo","doi":"10.1109/ICCV48922.2021.01503","DOIUrl":"https://doi.org/10.1109/ICCV48922.2021.01503","url":null,"abstract":"Data-driven sky models have gained much attention in outdoor illumination prediction recently, showing superior performance against analytical models. However, naively compressing an outdoor panorama into a low-dimensional latent vector, as existing models have done, causes two major problems. One is the mutual interference between the HDR intensity of the sun and the complex textures of the surrounding sky, and the other is the lack of fine-grained control over independent lighting factors due to the entangled representation. To address these issues, we propose a hierarchical disentangled sky model (HDSky) for outdoor illumination prediction. With this model, any outdoor panorama can be hierarchically disentangled into several factors based on three well-designed autoencoders. The first autoencoder compresses each sunny panorama into a sky vector and a sun vector with some constraints. The second autoencoder and the third autoencoder further disentangle the sun intensity and the sky intensity from the sun vector and the sky vector with several customized loss functions respectively. Moreover, a unified framework is designed to predict all-weather sky information from a single outdoor image. Through extensive experiments, we demonstrate that the proposed model significantly improves the accuracy of outdoor illumination prediction. It also allows users to intuitively edit the predicted panorama (e.g., changing the position of the sun while preserving others), without sacrificing physical plausibility.","PeriodicalId":6820,"journal":{"name":"2021 IEEE/CVF International Conference on Computer Vision (ICCV)","volume":"79 1","pages":"15293-15302"},"PeriodicalIF":0.0,"publicationDate":"2021-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"80359899","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-10-01DOI: 10.1109/ICCV48922.2021.00403
Ellen D. Zhong, Adam K. Lerer, Joey Davis, B. Berger
Protein structure determination from cryo-EM data requires reconstructing a 3D volume (or distribution of volumes) from many noisy and randomly oriented 2D projection images. While the standard homogeneous reconstruction task aims to recover a single static structure, recently-proposed neural and non-neural methods can reconstruct distributions of structures, thereby enabling the study of protein complexes that possess intrinsic structural or conformational heterogeneity. These heterogeneous reconstruction methods, however, require fixed image poses, which are typically estimated from an upstream homogeneous reconstruction and are not guaranteed to be accurate under highly heterogeneous conditions.In this work we describe cryoDRGN2, an ab initio reconstruction algorithm, which can jointly estimate image poses and learn a neural model of a distribution of 3D structures on real heterogeneous cryo-EM data. To achieve this, we adapt search algorithms from the traditional cryo-EM literature, and describe the optimizations and design choices required to make such a search procedure computationally tractable in the neural model setting. We show that cryoDRGN2 is robust to the high noise levels of real cryo-EM images, trains faster than earlier neural methods, and achieves state-of-the-art performance on real cryo-EM datasets.
{"title":"CryoDRGN2: Ab initio neural reconstruction of 3D protein structures from real cryo-EM images","authors":"Ellen D. Zhong, Adam K. Lerer, Joey Davis, B. Berger","doi":"10.1109/ICCV48922.2021.00403","DOIUrl":"https://doi.org/10.1109/ICCV48922.2021.00403","url":null,"abstract":"Protein structure determination from cryo-EM data requires reconstructing a 3D volume (or distribution of volumes) from many noisy and randomly oriented 2D projection images. While the standard homogeneous reconstruction task aims to recover a single static structure, recently-proposed neural and non-neural methods can reconstruct distributions of structures, thereby enabling the study of protein complexes that possess intrinsic structural or conformational heterogeneity. These heterogeneous reconstruction methods, however, require fixed image poses, which are typically estimated from an upstream homogeneous reconstruction and are not guaranteed to be accurate under highly heterogeneous conditions.In this work we describe cryoDRGN2, an ab initio reconstruction algorithm, which can jointly estimate image poses and learn a neural model of a distribution of 3D structures on real heterogeneous cryo-EM data. To achieve this, we adapt search algorithms from the traditional cryo-EM literature, and describe the optimizations and design choices required to make such a search procedure computationally tractable in the neural model setting. We show that cryoDRGN2 is robust to the high noise levels of real cryo-EM images, trains faster than earlier neural methods, and achieves state-of-the-art performance on real cryo-EM datasets.","PeriodicalId":6820,"journal":{"name":"2021 IEEE/CVF International Conference on Computer Vision (ICCV)","volume":"195 1","pages":"4046-4055"},"PeriodicalIF":0.0,"publicationDate":"2021-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"75884883","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
This work proposes a novel Deep Neural Network (DNN) quantization framework, namely RMSMP, with a Row-wise Mixed-Scheme and Multi-Precision approach. Specifically, this is the first effort to assign mixed quantization schemes and multiple precisions within layers – among rows of the DNN weight matrix, for simplified operations in hardware inference, while preserving accuracy. Furthermore, this paper makes a different observation from the prior work that the quantization error does not necessarily exhibit the layer-wise sensitivity, and actually can be mitigated as long as a certain portion of the weights in every layer are in higher precisions. This observation enables layer-wise uniformality in the hardware implementation towards guaranteed inference acceleration, while still enjoying row-wise flexibility of mixed schemes and multiple precisions to boost accuracy. The candidates of schemes and precisions are derived practically and effectively with a highly hardware-informative strategy to reduce the problem search space.With the offline determined ratio of different quantization schemes and precisions for all the layers, the RMSMP quantization algorithm uses Hessian and variance based method to effectively assign schemes and precisions for each row. The proposed RMSMP is tested for the image classification and natural language processing (BERT) applications, and achieves the best accuracy performance among state-of-the-arts under the same equivalent precisions. The RMSMP is implemented on FPGA devices, achieving 3.65× speedup in the end-to-end inference time for ResNet-18 on ImageNet, comparing with the 4-bit Fixed-point baseline.
{"title":"RMSMP: A Novel Deep Neural Network Quantization Framework with Row-wise Mixed Schemes and Multiple Precisions","authors":"Sung-En Chang, Yanyu Li, Mengshu Sun, Weiwen Jiang, Sijia Liu, Yanzhi Wang, Xue Lin","doi":"10.1109/ICCV48922.2021.00520","DOIUrl":"https://doi.org/10.1109/ICCV48922.2021.00520","url":null,"abstract":"This work proposes a novel Deep Neural Network (DNN) quantization framework, namely RMSMP, with a Row-wise Mixed-Scheme and Multi-Precision approach. Specifically, this is the first effort to assign mixed quantization schemes and multiple precisions within layers – among rows of the DNN weight matrix, for simplified operations in hardware inference, while preserving accuracy. Furthermore, this paper makes a different observation from the prior work that the quantization error does not necessarily exhibit the layer-wise sensitivity, and actually can be mitigated as long as a certain portion of the weights in every layer are in higher precisions. This observation enables layer-wise uniformality in the hardware implementation towards guaranteed inference acceleration, while still enjoying row-wise flexibility of mixed schemes and multiple precisions to boost accuracy. The candidates of schemes and precisions are derived practically and effectively with a highly hardware-informative strategy to reduce the problem search space.With the offline determined ratio of different quantization schemes and precisions for all the layers, the RMSMP quantization algorithm uses Hessian and variance based method to effectively assign schemes and precisions for each row. The proposed RMSMP is tested for the image classification and natural language processing (BERT) applications, and achieves the best accuracy performance among state-of-the-arts under the same equivalent precisions. The RMSMP is implemented on FPGA devices, achieving 3.65× speedup in the end-to-end inference time for ResNet-18 on ImageNet, comparing with the 4-bit Fixed-point baseline.","PeriodicalId":6820,"journal":{"name":"2021 IEEE/CVF International Conference on Computer Vision (ICCV)","volume":"25 1","pages":"5231-5240"},"PeriodicalIF":0.0,"publicationDate":"2021-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"82243688","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-10-01DOI: 10.1109/ICCV48922.2021.01318
Tianjiao Li, Qiuhong Ke, Hossein Rahmani, Rui En Ho, Henghui Ding, Jun Liu
Most of the state-of-the-art action recognition methods focus on offline learning, where the samples of all types of actions need to be provided at once. Here, we address continual learning of action recognition, where various types of new actions are continuously learned over time. This task is quite challenging, owing to the catastrophic forgetting problem stemming from the discrepancies between the previously learned actions and current new actions to be learned. Therefore, we propose Else-Net, a novel Elastic Semantic Network with multiple learning blocks to learn diversified human actions over time. Specifically, our Else-Net is able to automatically search and update the most relevant learning blocks w.r.t. the current new action, or explore new blocks to store new knowledge, preserving the unmatched ones to retain the knowledge of previously learned actions and alleviates forgetting when learning new actions. Moreover, even though different human actions may vary to a large extent as a whole, their local body parts can still share many homogeneous features. Inspired by this, our proposed Else-Net mines the shared knowledge of the decomposed human body parts from different actions, which benefits continual learning of actions. Experiments show that the proposed approach enables effective continual action recognition and achieves promising performance on two large-scale action recognition datasets.
{"title":"Else-Net: Elastic Semantic Network for Continual Action Recognition from Skeleton Data","authors":"Tianjiao Li, Qiuhong Ke, Hossein Rahmani, Rui En Ho, Henghui Ding, Jun Liu","doi":"10.1109/ICCV48922.2021.01318","DOIUrl":"https://doi.org/10.1109/ICCV48922.2021.01318","url":null,"abstract":"Most of the state-of-the-art action recognition methods focus on offline learning, where the samples of all types of actions need to be provided at once. Here, we address continual learning of action recognition, where various types of new actions are continuously learned over time. This task is quite challenging, owing to the catastrophic forgetting problem stemming from the discrepancies between the previously learned actions and current new actions to be learned. Therefore, we propose Else-Net, a novel Elastic Semantic Network with multiple learning blocks to learn diversified human actions over time. Specifically, our Else-Net is able to automatically search and update the most relevant learning blocks w.r.t. the current new action, or explore new blocks to store new knowledge, preserving the unmatched ones to retain the knowledge of previously learned actions and alleviates forgetting when learning new actions. Moreover, even though different human actions may vary to a large extent as a whole, their local body parts can still share many homogeneous features. Inspired by this, our proposed Else-Net mines the shared knowledge of the decomposed human body parts from different actions, which benefits continual learning of actions. Experiments show that the proposed approach enables effective continual action recognition and achieves promising performance on two large-scale action recognition datasets.","PeriodicalId":6820,"journal":{"name":"2021 IEEE/CVF International Conference on Computer Vision (ICCV)","volume":"36 1","pages":"13414-13423"},"PeriodicalIF":0.0,"publicationDate":"2021-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"82246088","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-10-01DOI: 10.1109/ICCV48922.2021.01331
Mang Ye, Weijian Ruan, Bo Du, Mike Zheng Shou
This paper introduces a powerful channel augmented joint learning strategy for the visible-infrared recognition problem. For data augmentation, most existing methods directly adopt the standard operations designed for single-modality visible images, and thus do not fully consider the imagery properties in visible to infrared matching. Our basic idea is to homogenously generate color-irrelevant images by randomly exchanging the color channels. It can be seamlessly integrated into existing augmentation operations without modifying the network, consistently improving the robustness against color variations. Incorporated with a random erasing strategy, it further greatly enriches the diversity by simulating random occlusions. For cross-modality metric learning, we design an enhanced channel-mixed learning strategy to simultaneously handle the intra-and cross-modality variations with squared difference for stronger discriminability. Besides, a channel-augmented joint learning strategy is further developed to explicitly optimize the outputs of augmented images. Extensive experiments with insightful analysis on two visible-infrared recognition tasks show that the proposed strategies consistently improve the accuracy. Without auxiliary information, it improves the state-of-the-art Rank-1/mAP by 14.59%/13.00% on the large-scale SYSU-MM01 dataset.
{"title":"Channel Augmented Joint Learning for Visible-Infrared Recognition","authors":"Mang Ye, Weijian Ruan, Bo Du, Mike Zheng Shou","doi":"10.1109/ICCV48922.2021.01331","DOIUrl":"https://doi.org/10.1109/ICCV48922.2021.01331","url":null,"abstract":"This paper introduces a powerful channel augmented joint learning strategy for the visible-infrared recognition problem. For data augmentation, most existing methods directly adopt the standard operations designed for single-modality visible images, and thus do not fully consider the imagery properties in visible to infrared matching. Our basic idea is to homogenously generate color-irrelevant images by randomly exchanging the color channels. It can be seamlessly integrated into existing augmentation operations without modifying the network, consistently improving the robustness against color variations. Incorporated with a random erasing strategy, it further greatly enriches the diversity by simulating random occlusions. For cross-modality metric learning, we design an enhanced channel-mixed learning strategy to simultaneously handle the intra-and cross-modality variations with squared difference for stronger discriminability. Besides, a channel-augmented joint learning strategy is further developed to explicitly optimize the outputs of augmented images. Extensive experiments with insightful analysis on two visible-infrared recognition tasks show that the proposed strategies consistently improve the accuracy. Without auxiliary information, it improves the state-of-the-art Rank-1/mAP by 14.59%/13.00% on the large-scale SYSU-MM01 dataset.","PeriodicalId":6820,"journal":{"name":"2021 IEEE/CVF International Conference on Computer Vision (ICCV)","volume":"46 1","pages":"13547-13556"},"PeriodicalIF":0.0,"publicationDate":"2021-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"81160858","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-10-01DOI: 10.1109/ICCV48922.2021.00416
Wei-Ting Chen, H. Fang, Cheng-Lin Hsieh, Cheng-Che Tsai, I-Hsiang Chen, Jianwei Ding, Sy-Yen Kuo
Snow is a highly complicated atmospheric phenomenon that usually contains snowflake, snow streak, and veiling effect (similar to the haze or the mist). In this literature, we propose a single image desnowing algorithm to address the diversity of snow particles in shape and size. First, to better represent the complex snow shape, we apply the dual-tree wavelet transform and propose a complex wavelet loss in the network. Second, we propose a hierarchical decomposition paradigm in our network for better under-standing the different sizes of snow particles. Last, we propose a novel feature called the contradict channel (CC) for the snow scenes. We find that the regions containing the snow particles tend to have higher intensity in the CC than that in the snow-free regions. We leverage this discriminative feature to construct the contradict channel loss for improving the performance of snow removal. Moreover, due to the limitation of existing snow datasets, to simulate the snow scenarios comprehensively, we propose a large-scale dataset called Comprehensive Snow Dataset (CSD). Experimental results show that the proposed method can favorably outperform existing methods in three synthetic datasets and real-world datasets. The code and dataset are released in https://github.com/weitingchen83/ICCV2021-Single-Image-Desnowing-HDCWNet.
{"title":"ALL Snow Removed: Single Image Desnowing Algorithm Using Hierarchical Dual-tree Complex Wavelet Representation and Contradict Channel Loss","authors":"Wei-Ting Chen, H. Fang, Cheng-Lin Hsieh, Cheng-Che Tsai, I-Hsiang Chen, Jianwei Ding, Sy-Yen Kuo","doi":"10.1109/ICCV48922.2021.00416","DOIUrl":"https://doi.org/10.1109/ICCV48922.2021.00416","url":null,"abstract":"Snow is a highly complicated atmospheric phenomenon that usually contains snowflake, snow streak, and veiling effect (similar to the haze or the mist). In this literature, we propose a single image desnowing algorithm to address the diversity of snow particles in shape and size. First, to better represent the complex snow shape, we apply the dual-tree wavelet transform and propose a complex wavelet loss in the network. Second, we propose a hierarchical decomposition paradigm in our network for better under-standing the different sizes of snow particles. Last, we propose a novel feature called the contradict channel (CC) for the snow scenes. We find that the regions containing the snow particles tend to have higher intensity in the CC than that in the snow-free regions. We leverage this discriminative feature to construct the contradict channel loss for improving the performance of snow removal. Moreover, due to the limitation of existing snow datasets, to simulate the snow scenarios comprehensively, we propose a large-scale dataset called Comprehensive Snow Dataset (CSD). Experimental results show that the proposed method can favorably outperform existing methods in three synthetic datasets and real-world datasets. The code and dataset are released in https://github.com/weitingchen83/ICCV2021-Single-Image-Desnowing-HDCWNet.","PeriodicalId":6820,"journal":{"name":"2021 IEEE/CVF International Conference on Computer Vision (ICCV)","volume":"32 1","pages":"4176-4185"},"PeriodicalIF":0.0,"publicationDate":"2021-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"89800029","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：