Pub Date : 2020-12-01DOI: 10.1109/VCIP49819.2020.9301861
Nakamasa Inoue
This paper proposes Graph Grouping (GG) loss for metric learning and its application to face verification. GG loss predisposes image embeddings of the same identity to be close to each other, and those of different identities to be far from each other by constructing and optimizing graphs representing the relation between images. Further, to reduce the computational cost, we propose an efficient way to compute GG loss for cases where embeddings are L2 normalized. In experiments, we demonstrate the effectiveness of the proposed method for face verification on the VoxCeleb dataset. The results show that the proposed GG loss outperforms conventional losses for metric learning.
{"title":"Graph Grouping Loss for Metric Learning of Face Image Representations","authors":"Nakamasa Inoue","doi":"10.1109/VCIP49819.2020.9301861","DOIUrl":"https://doi.org/10.1109/VCIP49819.2020.9301861","url":null,"abstract":"This paper proposes Graph Grouping (GG) loss for metric learning and its application to face verification. GG loss predisposes image embeddings of the same identity to be close to each other, and those of different identities to be far from each other by constructing and optimizing graphs representing the relation between images. Further, to reduce the computational cost, we propose an efficient way to compute GG loss for cases where embeddings are L2 normalized. In experiments, we demonstrate the effectiveness of the proposed method for face verification on the VoxCeleb dataset. The results show that the proposed GG loss outperforms conventional losses for metric learning.","PeriodicalId":431880,"journal":{"name":"2020 IEEE International Conference on Visual Communications and Image Processing (VCIP)","volume":"16 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128142040","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 : 2020-12-01DOI: 10.1109/VCIP49819.2020.9301857
Z. Chen, Hantao Wang, Lijun Wu, Yanlin Zhou, Dapeng Oliver Wu
Depth completion aims to estimate dense depth maps from sparse depth measurements. It has become increasingly important in autonomous driving and thus has drawn wide attention. In this paper, we introduce photometric losses in both spatial and time domains to jointly guide self-supervised depth completion. This method performs an accurate end-to-end depth completion of vision tasks by using LiDAR and a monocular camera. In particular, we full utilize the consistent information inside the temporally adjacent frames and the stereo vision to improve the accuracy of depth completion in the model training phase. We design a self-supervised framework to eliminate the negative effects of moving objects and the region with smooth gradients. Experiments are conducted on KITTI. Results indicate that our self-supervised method can attain competitive performance.
{"title":"Spatiotemporal Guided Self-Supervised Depth Completion from LiDAR and Monocular Camera","authors":"Z. Chen, Hantao Wang, Lijun Wu, Yanlin Zhou, Dapeng Oliver Wu","doi":"10.1109/VCIP49819.2020.9301857","DOIUrl":"https://doi.org/10.1109/VCIP49819.2020.9301857","url":null,"abstract":"Depth completion aims to estimate dense depth maps from sparse depth measurements. It has become increasingly important in autonomous driving and thus has drawn wide attention. In this paper, we introduce photometric losses in both spatial and time domains to jointly guide self-supervised depth completion. This method performs an accurate end-to-end depth completion of vision tasks by using LiDAR and a monocular camera. In particular, we full utilize the consistent information inside the temporally adjacent frames and the stereo vision to improve the accuracy of depth completion in the model training phase. We design a self-supervised framework to eliminate the negative effects of moving objects and the region with smooth gradients. Experiments are conducted on KITTI. Results indicate that our self-supervised method can attain competitive performance.","PeriodicalId":431880,"journal":{"name":"2020 IEEE International Conference on Visual Communications and Image Processing (VCIP)","volume":"26 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128686063","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 : 2020-12-01DOI: 10.1109/VCIP49819.2020.9301790
Kun-Min Yang, Dong Liu, Feng Wu
In the hybrid video coding framework, transform is adopted to exploit the dependency within the input signal. In this paper, we propose a deep learning-based nonlinear transform for intra coding. Specifically, we incorporate the directional information into the residual domain. Then, a convolutional neural network model is designed to achieve better decorrelation and energy compaction than the conventional discrete cosine transform. This work has two main contributions. First, we propose to use the intra prediction signal to reduce the directionality in the residual. Second, we present a novel loss function to characterize the efficiency of the transform during the training. To evaluate the compression performance of the proposed transform, we implement it into the High Efficiency Video Coding reference software. Experimental results demonstrate that the proposed method achieves up to 1.79% BD-rate reduction for natural videos.
{"title":"Deep Learning-Based Nonlinear Transform for HEVC Intra Coding","authors":"Kun-Min Yang, Dong Liu, Feng Wu","doi":"10.1109/VCIP49819.2020.9301790","DOIUrl":"https://doi.org/10.1109/VCIP49819.2020.9301790","url":null,"abstract":"In the hybrid video coding framework, transform is adopted to exploit the dependency within the input signal. In this paper, we propose a deep learning-based nonlinear transform for intra coding. Specifically, we incorporate the directional information into the residual domain. Then, a convolutional neural network model is designed to achieve better decorrelation and energy compaction than the conventional discrete cosine transform. This work has two main contributions. First, we propose to use the intra prediction signal to reduce the directionality in the residual. Second, we present a novel loss function to characterize the efficiency of the transform during the training. To evaluate the compression performance of the proposed transform, we implement it into the High Efficiency Video Coding reference software. Experimental results demonstrate that the proposed method achieves up to 1.79% BD-rate reduction for natural videos.","PeriodicalId":431880,"journal":{"name":"2020 IEEE International Conference on Visual Communications and Image Processing (VCIP)","volume":"28 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132902787","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 : 2020-12-01DOI: 10.1109/VCIP49819.2020.9301879
Bernardo Beling, Iago Storch, L. Agostini, B. Zatt, S. Bampi, D. Palomino
This work presents an Equirectangular projection (ERP) based Coding Tree Unit (CTU) splitting early termination algorithm for the High Efficiency Video Coding (HEVC) intra prediction of 360-degree videos. The proposed algorithm adaptively employs early termination in the HEVC CTU splitting based on distortion properties of the ERP projection, that generate homogeneous regions at the top and bottom portion of a video frame. Experimental results show an average of 24% time saving with 0.11% coding efficiency loss, significantly reducing the encoding complexity with minor impacts in the encoding efficiency. Besides, solution presents the best results considering the relation between time saving and coding efficiency when compared with all related works.
{"title":"ERP-Based CTU Splitting Early Termination for Intra Prediction of 360 videos","authors":"Bernardo Beling, Iago Storch, L. Agostini, B. Zatt, S. Bampi, D. Palomino","doi":"10.1109/VCIP49819.2020.9301879","DOIUrl":"https://doi.org/10.1109/VCIP49819.2020.9301879","url":null,"abstract":"This work presents an Equirectangular projection (ERP) based Coding Tree Unit (CTU) splitting early termination algorithm for the High Efficiency Video Coding (HEVC) intra prediction of 360-degree videos. The proposed algorithm adaptively employs early termination in the HEVC CTU splitting based on distortion properties of the ERP projection, that generate homogeneous regions at the top and bottom portion of a video frame. Experimental results show an average of 24% time saving with 0.11% coding efficiency loss, significantly reducing the encoding complexity with minor impacts in the encoding efficiency. Besides, solution presents the best results considering the relation between time saving and coding efficiency when compared with all related works.","PeriodicalId":431880,"journal":{"name":"2020 IEEE International Conference on Visual Communications and Image Processing (VCIP)","volume":"75 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127806240","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 : 2020-12-01DOI: 10.1109/VCIP49819.2020.9301878
C. Kok, Wing-Shan Tam
The tutorial starts with an introduction of digital image interpolation, and single image super-resolution. It continues with the definition of various image interpolation performance measurement indices, including both objective and subjective indices. The core of this tutorial is the application of covariance based interpolation to achieve high visual quality image interpolation and single image super-resolution results. Layer on layer, the covariance based edge-directed image interpolation techniques that makes use of stochastic image model without explicit edge map, to iterative covariance correction based image interpolation. The edge based interpolation incorporated human visual system to achieve visually pleasant high resolution interpolation results. On each layer, the pros and cons of each image model and interpolation technique, solutions to alleviate the interpolation visual artifacts of each techniques, and innovative modification to overcome limitations of traditional edge-directed image interpolation techniques are presented in this tutorial, which includes: spatial adaptive pixel intensity estimation, pixel intensity correction, error propagation mitigation, covariance windows adaptation, and iterative covariance correction. The tutorial will extend from theoretical and analytical discussions to detail implementation using MATLAB. The audience shall be able to bring home with implementation details, as well as the performance and complexity of the interpolation algorithms discussed in this tutorial.
{"title":"From Low to Super Resolution and Beyond","authors":"C. Kok, Wing-Shan Tam","doi":"10.1109/VCIP49819.2020.9301878","DOIUrl":"https://doi.org/10.1109/VCIP49819.2020.9301878","url":null,"abstract":"The tutorial starts with an introduction of digital image interpolation, and single image super-resolution. It continues with the definition of various image interpolation performance measurement indices, including both objective and subjective indices. The core of this tutorial is the application of covariance based interpolation to achieve high visual quality image interpolation and single image super-resolution results. Layer on layer, the covariance based edge-directed image interpolation techniques that makes use of stochastic image model without explicit edge map, to iterative covariance correction based image interpolation. The edge based interpolation incorporated human visual system to achieve visually pleasant high resolution interpolation results. On each layer, the pros and cons of each image model and interpolation technique, solutions to alleviate the interpolation visual artifacts of each techniques, and innovative modification to overcome limitations of traditional edge-directed image interpolation techniques are presented in this tutorial, which includes: spatial adaptive pixel intensity estimation, pixel intensity correction, error propagation mitigation, covariance windows adaptation, and iterative covariance correction. The tutorial will extend from theoretical and analytical discussions to detail implementation using MATLAB. The audience shall be able to bring home with implementation details, as well as the performance and complexity of the interpolation algorithms discussed in this tutorial.","PeriodicalId":431880,"journal":{"name":"2020 IEEE International Conference on Visual Communications and Image Processing (VCIP)","volume":"8 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131867500","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 : 2020-12-01DOI: 10.1109/VCIP49819.2020.9301818
Yuge Zhang, Min Zhao, Longbin Yan, Tiande Gao, Jie Chen
Recently, face recognition systems have received significant attention, and there have been many works focused on presentation attacks (PAs). However, the generalization capacity of PAs is still challenging in real scenarios, as the attack samples in the training database may not cover all possible PAs. In this paper, we propose to perform the face presentation attack detection (PAD) with multi-channel images using the convolutional neural network based anomaly detection. Multi-channel images endow us with rich information to distinguish between different mode of attacks, and the anomaly detection based technique ensures the generalization performance. We evaluate the performance of our methods using the wide multi-channel presentation attack (WMCA) dataset.
{"title":"CNN-Based Anomaly Detection For Face Presentation Attack Detection With Multi-Channel Images","authors":"Yuge Zhang, Min Zhao, Longbin Yan, Tiande Gao, Jie Chen","doi":"10.1109/VCIP49819.2020.9301818","DOIUrl":"https://doi.org/10.1109/VCIP49819.2020.9301818","url":null,"abstract":"Recently, face recognition systems have received significant attention, and there have been many works focused on presentation attacks (PAs). However, the generalization capacity of PAs is still challenging in real scenarios, as the attack samples in the training database may not cover all possible PAs. In this paper, we propose to perform the face presentation attack detection (PAD) with multi-channel images using the convolutional neural network based anomaly detection. Multi-channel images endow us with rich information to distinguish between different mode of attacks, and the anomaly detection based technique ensures the generalization performance. We evaluate the performance of our methods using the wide multi-channel presentation attack (WMCA) dataset.","PeriodicalId":431880,"journal":{"name":"2020 IEEE International Conference on Visual Communications and Image Processing (VCIP)","volume":"223 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134235938","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 : 2020-12-01DOI: 10.1109/VCIP49819.2020.9301755
Teli Ma, Yizhi Wang, Jinxin Shao, Baochang Zhang, D. Doermann
Generative models have been successfully used for anomaly detection, which however need a large number of parameters and computation overheads, especially when training spatial and temporal networks in the same framework. In this paper, we introduce a novel network architecture, Orthogonal Features Fusion Network (OFF-Net), to solve the anomaly detection problem. We show that the convolutional feature maps used for generating future frames are orthogonal with each other, which can improve representation capacity of generative models and strengthen temporal connections between adjacent images. We lead a simple but effective module easily mounted on convolutional neural networks (CNNs) with negligible additional parameters added, which can replace the widely-used optical flow n etwork a nd s ignificantly im prove th e pe rformance for anomaly detection. Extensive experiment results demonstrate the effectiveness of OFF-Net that we outperform the state-of-the-art model 1.7% in terms of AUC. We save around 85M-space parameters compared with the prevailing prior arts using optical flow n etwork w ithout c omprising t he performance.
{"title":"Orthogonal Features Fusion Network for Anomaly Detection","authors":"Teli Ma, Yizhi Wang, Jinxin Shao, Baochang Zhang, D. Doermann","doi":"10.1109/VCIP49819.2020.9301755","DOIUrl":"https://doi.org/10.1109/VCIP49819.2020.9301755","url":null,"abstract":"Generative models have been successfully used for anomaly detection, which however need a large number of parameters and computation overheads, especially when training spatial and temporal networks in the same framework. In this paper, we introduce a novel network architecture, Orthogonal Features Fusion Network (OFF-Net), to solve the anomaly detection problem. We show that the convolutional feature maps used for generating future frames are orthogonal with each other, which can improve representation capacity of generative models and strengthen temporal connections between adjacent images. We lead a simple but effective module easily mounted on convolutional neural networks (CNNs) with negligible additional parameters added, which can replace the widely-used optical flow n etwork a nd s ignificantly im prove th e pe rformance for anomaly detection. Extensive experiment results demonstrate the effectiveness of OFF-Net that we outperform the state-of-the-art model 1.7% in terms of AUC. We save around 85M-space parameters compared with the prevailing prior arts using optical flow n etwork w ithout c omprising t he performance.","PeriodicalId":431880,"journal":{"name":"2020 IEEE International Conference on Visual Communications and Image Processing (VCIP)","volume":"13 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129733259","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 : 2020-12-01DOI: 10.1109/VCIP49819.2020.9301892
Xiuyuan Wang, Yikun Pan, D. Lun
Reflection removal is a long-standing problem in computer vision. In this paper, we consider the reflection removal problem for stereoscopic images. By exploiting the depth information of stereoscopic images, a new background edge estimation algorithm based on the Wasserstein Generative Adversarial Network (WGAN) is proposed to distinguish the edges of the background image from the reflection. The background edges are then used to reconstruct the background image. We compare the proposed approach with the state-of-the- art reflection removal methods. Results show that the proposed approach can outperform the traditional single-image based methods and is comparable to the multiple-image based approach while having a much simpler imaging hardware requirement.
{"title":"Stereoscopic image reflection removal based on Wasserstein Generative Adversarial Network","authors":"Xiuyuan Wang, Yikun Pan, D. Lun","doi":"10.1109/VCIP49819.2020.9301892","DOIUrl":"https://doi.org/10.1109/VCIP49819.2020.9301892","url":null,"abstract":"Reflection removal is a long-standing problem in computer vision. In this paper, we consider the reflection removal problem for stereoscopic images. By exploiting the depth information of stereoscopic images, a new background edge estimation algorithm based on the Wasserstein Generative Adversarial Network (WGAN) is proposed to distinguish the edges of the background image from the reflection. The background edges are then used to reconstruct the background image. We compare the proposed approach with the state-of-the- art reflection removal methods. Results show that the proposed approach can outperform the traditional single-image based methods and is comparable to the multiple-image based approach while having a much simpler imaging hardware requirement.","PeriodicalId":431880,"journal":{"name":"2020 IEEE International Conference on Visual Communications and Image Processing (VCIP)","volume":"67 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129436681","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 : 2020-12-01DOI: 10.1109/VCIP49819.2020.9301773
Jinzhao Zhou, Xingming Zhang, Yang Liu
Previous studies recognizing expressions with facial graph topology mostly use a fixed facial graph structure established by the physical dependencies among facial landmarks. However, the static graph structure inherently lacks flexibility in non-standardized scenarios. This paper proposes a dynamic-graph-based method for effective and robust facial expression recognition. To capture action-specific dependencies among facial components, we introduce a link inference structure, called the Situational Link Generation Module (SLGM). We further propose the Situational Graph Convolution Network (SGCN) to automatically detect and recognize facial expression in various conditions. Experimental evaluations on two lab-constrained datasets, CK+ and Oulu, along with an in-the-wild dataset, AFEW, show the superior performance of the proposed method. Additional experiments on occluded facial images further demonstrate the robustness of our strategy.
{"title":"Learning the Connectivity: Situational Graph Convolution Network for Facial Expression Recognition","authors":"Jinzhao Zhou, Xingming Zhang, Yang Liu","doi":"10.1109/VCIP49819.2020.9301773","DOIUrl":"https://doi.org/10.1109/VCIP49819.2020.9301773","url":null,"abstract":"Previous studies recognizing expressions with facial graph topology mostly use a fixed facial graph structure established by the physical dependencies among facial landmarks. However, the static graph structure inherently lacks flexibility in non-standardized scenarios. This paper proposes a dynamic-graph-based method for effective and robust facial expression recognition. To capture action-specific dependencies among facial components, we introduce a link inference structure, called the Situational Link Generation Module (SLGM). We further propose the Situational Graph Convolution Network (SGCN) to automatically detect and recognize facial expression in various conditions. Experimental evaluations on two lab-constrained datasets, CK+ and Oulu, along with an in-the-wild dataset, AFEW, show the superior performance of the proposed method. Additional experiments on occluded facial images further demonstrate the robustness of our strategy.","PeriodicalId":431880,"journal":{"name":"2020 IEEE International Conference on Visual Communications and Image Processing (VCIP)","volume":"13 3-4 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122345591","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 : 2020-12-01DOI: 10.1109/VCIP49819.2020.9301848
Chaoqun Lin, R. Guo, Mingkun Li, Xianbiao Qi, Chun-Guang Li
Person Re-Identification (Re-ID) is a challenging task of matching pedestrian images collected from nonoverlapping multiple camera views due to huge variations from pose changes, occlusions, varying illumination and clutter background. Recently, graph convolution network or graph neural network increasingly gains a lot of research attention in person Re-ID. However, the existing methods have not fully exploit the available features on the graph. In this paper, we propose an efficient and effective end-to-end trainable framework, termed Edge Attention Convolution Network (EACN), to perform convolution feature learning and attentive feature aggregation for person Re-ID, in which the learned convolution features on vertex and its edges are attentively aggregated on a dynamic graph. We conduct extensive experiments on two large benchmark datasets, Market-1501 and DukeMTMC. Experimental results validate the efficiency and effectiveness of our proposal.
{"title":"Learning Convolution Feature Aggregation via Edge Attention Convolution Network for Person Re-Identification","authors":"Chaoqun Lin, R. Guo, Mingkun Li, Xianbiao Qi, Chun-Guang Li","doi":"10.1109/VCIP49819.2020.9301848","DOIUrl":"https://doi.org/10.1109/VCIP49819.2020.9301848","url":null,"abstract":"Person Re-Identification (Re-ID) is a challenging task of matching pedestrian images collected from nonoverlapping multiple camera views due to huge variations from pose changes, occlusions, varying illumination and clutter background. Recently, graph convolution network or graph neural network increasingly gains a lot of research attention in person Re-ID. However, the existing methods have not fully exploit the available features on the graph. In this paper, we propose an efficient and effective end-to-end trainable framework, termed Edge Attention Convolution Network (EACN), to perform convolution feature learning and attentive feature aggregation for person Re-ID, in which the learned convolution features on vertex and its edges are attentively aggregated on a dynamic graph. We conduct extensive experiments on two large benchmark datasets, Market-1501 and DukeMTMC. Experimental results validate the efficiency and effectiveness of our proposal.","PeriodicalId":431880,"journal":{"name":"2020 IEEE International Conference on Visual Communications and Image Processing (VCIP)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128983936","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：