Adversarially Occluded Samples for Person Re-identification

Huang, Houjing; Li, Dangwei; Zhang, Zhang; Chen, Xiaotang; Huang, Kaiqi

doi:10.1109/cvpr.2018.00535

Cited by 235 publications

(105 citation statements)

References 31 publications

Supporting

Mentioning

105

Contrasting

Order By: Relevance

“…Pyramid-ours 79.0 89.0 MGN [26] 78.4 88.7 SVDNet [23] 56.8 76.7 AOS [8] 62.1 79.2 HA-CNN [14] 63.8 80.5 GSRW [20] 66.4 80.7 DuATM [21] 64.6 81.8 PCB+RPP [24] 69.2 83.3 PSE+ECN [19] 75.7 84.5 DNN-CRF [3] 69.5 84.9 GP-reid [28] 72.8 85.2 Table 3. Comparison results (%) on CUHK03 dataset using the new protocol in [38].…”

Section: Comparison With State-of-the-art Methodsmentioning

confidence: 99%

See 1 more Smart Citation

Pyramidal Person Re-IDentification via Multi-Loss Dynamic Training

Zheng

Deng

Sun

et al. 2019

2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)

326

202

View full text Add to dashboard Cite

Most existing Re-IDentification (Re-ID) methods are highly dependent on precise bounding boxes that enable images to be aligned with each other. However, due to the challenging practical scenarios, current detection models often produce inaccurate bounding boxes, which inevitably degenerate the performance of existing Re-ID algorithms. In this paper, we propose a novel coarse-to-fine pyramid model to relax the need of bounding boxes, which not only incorporates local and global information, but also integrates the gradual cues between them. The pyramid model is able to match at different scales and then search for the correct image of the same identity, even when the image pairs are not aligned. In addition, in order to learn discriminative identity representation, we explore a dynamic training scheme to seamlessly unify two losses and extract appropriate shared information between them. Experimental results clearly demonstrate that the proposed method achieves the state-of-the-art results on three datasets. Especially, our approach exceeds the current best method by 9.5% on the most challenging CUHK03 dataset.

show abstract

Section: Comparison With State-of-the-art Methodsmentioning

confidence: 99%

“…Non part-based methods: Recently, a completely synthetic dataset [1] and some adversarially occluded samples [8] are constructed to train the re-identification model. In Figure 2.…”

Section: Related Workmentioning

confidence: 99%

Pyramidal Person Re-IDentification via Multi-Loss Dynamic Training

Zheng

Deng

Sun

et al. 2019

2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)

326

202

View full text Add to dashboard Cite

show abstract

“…Liu et al [40] propose the Adversarial Binary Coding (ABC) to guide the extraction of binary codes using adversarial learning for efficient person re-identification. Huang et al [41] adversarially generate occluded samples and combine them with training samples to fine-tune the CNN model.…”

Section: Adversarial Learningmentioning

confidence: 99%

Adversarial Erasing Attention for Person Re-Identification in Camera Networks Under Complex Environments

2020

View full text Add to dashboard Cite

Person re-identification (Re-ID) in camera networks under complex environments has achieved promising performance using deep feature representations. However, most approaches usually ignore to learn features from non-salient parts of pedestrian, which results in an incomplete pedestrian representation. In this paper, we propose a novel person Re-ID method named Adversarial Erasing Attention (AEA) to mine discriminative completed features using an adversarial way. Specifically, the proposed AEA consists of the basic network and the complementary network. On the one hand, original pedestrian images are used to train the basic network in order to extract global and local deep features. On the other hand, to learn features complementary to the basic network, we propose the adversarial erasing operation, that locates non-salient areas with the help of attention map, to generate erased pedestrian images. Then, we utilize them to train the complementary network and adopt the dynamic strategy to match the dynamic status of AEA in the learning process. Hence, the diversity of training samples is enriched and the complementary network could discover new clues when learning deep features. Finally, we combine the features learned from the basic and complementary networks to represent the pedestrian image. Experiments on three databases (Market1501, CUHK03 and DukeMTMC-reID) demonstrate the proposed AEA achieves great performances. INDEX TERMS Person re-identification, dynamic strategy, adversarial learning. I. INTRODUCTION Person re-identification (Re-ID) in camera networks targets to retrieve a specific pedestrian from a large gallery which is captured by multiple visual sensors [1]-[3]. There are many applications (such as multi-camera tracking [4], [5], crowd counting [6], [7] and so on [8], [9]) that require an accurate person Re-ID algorithm. However, it is a challenging task under complex environments because of significant changes in body poses, viewpoints, illuminations, etc. Recently, many researchers [10]-[14] employ Convolutional Neural Networks (CNNs) to extract pedestrian The associate editor coordinating the review of this manuscript and approving it for publication was Qilian Liang .

show abstract

“…Holistic Features Based Methods Given a backbone C-NN such as ResNet-50 [21] or other network architectures [2,51,71,78], this type of methods learns discriminative holistic features from the feature map directly. Specifically, they aim to learn the features by improving loss functions [9,14,22,31,41,42,50,55,63], improving the training techniques [1,4,12,24,32,35,37,54], adding additional network modules [23,23,51,62], using extra semantic annotations [30,46,47,79] or generating more training samples [17,33,72,76,77]. Besides, more recent studies [3,6,8,10,27,28,38,46,48,53,58,61,64,…”

Section: Related Workmentioning

confidence: 99%

Discriminative Spatial Feature Learning for Person Re-Identification

Peng

Huang

Wang

et al. 2020

Proceedings of the 28th ACM International Conference on Multimedia

View full text Add to dashboard Cite

Person re-identification (ReID) aims to match detected pedestrian images from multiple non-overlapping cameras. Most existing methods employ a backbone CNN to extract a vectorized feature representation by performing some global pooling operations (such as global average pooling and global max pooling) on the 3D feature map (i.e., the output of the backbone CNN). Although simple and effective in some situations, the global pooling operation only focuses on the statistical properties and ignores the spatial distribution of the feature map. Hence, it can not distinguish two feature maps when they have similar response values located in totally different positions. To handle this challenge, a novel method is proposed to learn the discriminative spatial features. Firstly, a self-constrained spatial transformer network (SC-STN) is introduced to handle the misalignments caused by detection errors. Then, based on the prior knowledge that the spatial structure of a pedestrian often keeps robust in vertical orientation of images, a novel vertical convolution network (VCN) is proposed to extract the spatial feature in vertical. Extensive experimental evaluations on several benchmarks demonstrate that the proposed method achieves state-of-theart performances by introducing only a few parameters to the backbone.

show abstract

Adversarially Occluded Samples for Person Re-identification

Cited by 235 publications

References 31 publications

Pyramidal Person Re-IDentification via Multi-Loss Dynamic Training

Pyramidal Person Re-IDentification via Multi-Loss Dynamic Training

Adversarial Erasing Attention for Person Re-Identification in Camera Networks Under Complex Environments

Discriminative Spatial Feature Learning for Person Re-Identification

Contact Info

Product

Resources

About