Cloth-Changing Person Re-identification from A Single Image with Gait Prediction and Regularization

Jin, Xin; He, Tianyu; Zheng, Kecheng; Yin, Zhiheng; Shen, Xu; Huang, Zhen; Feng, Ruoyu; Huang, Jianqiang; Hua, Xian-Sheng; Chen, Zhibo

doi:10.1109/cvpr52688.2022.01388

Cited by 50 publications

(16 citation statements)

References 50 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Since there are few works to explore CCVReID task, we compare our method with four kinds of methods for a comprehensive evaluation: 1) video-based person Re-ID methods that do not involve clothes-changing, including AP3D [17], TCLNet [22], and SINet [1], 2) gait recognition methods, including GaitSet [3], GaitPart [13], and GaitGL [29], 3) image-based clothes-changing methods, including ReIDCaps [23], Pixel sampling [37], and GI-ReID [24], 4) video-based clothes-changing method CAL [16]. Note that image-based methods receive a single frame as input, we report the results under two different settings of such methods, i.e., randomly select a frame for feature extraction (denoted with suffix "-R"), and extract the features of all frames and take the average as the final feature (denoted with suffix "-A").…”

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

confidence: 99%

“…[28] proposed a Clothing Agnostic Shape Extraction Network (CASE-Net) to shape-based feature representation via adversarial learning and feature disentanglement. Besides, other works attempted to leverage contour sketch [44], silhouettes [21,24], face [40], skeletons [34], 3D shape [6], or radio signals [14] to capture clothes-irrelevant features. Despite the achievement of image-based methods, they are susceptible to the quality of person images, i.e., they are less tolerant to noise due to the limited information contained in a single frame.…”

Section: Related Workmentioning

confidence: 99%

“…To solve that, [31,47] attempted to introduce gait to achieve complementarity of appearance and gait features. [24] performed gait prediction and regularization from a single image, which used gait to drive the main Re-ID model to learn cloth-independent features. However, such methods only consider the fusion of body and face information in a single sample.…”

Section: Related Workmentioning

confidence: 99%

See 2 more Smart Citations

A benchmark for clothes variation in person re‐identification

Wang

Chen

et al. 2020

Int J Intell Syst

View full text Add to dashboard Cite

Person re-identification (re-ID) has drawn attention significantly in the computer vision society due to its application and research significance. It aims to retrieve a person of interest across different camera views. However, there are still several factors that hinder the applications of person re-ID. In fact, most common data sets either assume that pedestrians do not change their clothing across different camera views or are taken under constrained environments. Those constraints simplify the person re-ID task and contribute to early development of person re-ID, yet a person has a great possibility to change clothes in real life. To facilitate the research toward conquering those issues, this paper mainly introduces a new benchmark data set for person reidentification. To the best of our knowledge, this data set is currently the most diverse for person re-identification. It contains 107 persons with 9,738 images, captured in 15 indoor/outdoor scenes from September 2019 to December 2019, varying according to viewpoints, lighting, resolutions, human pose, seasons, backgrounds, and clothes especially. We hope that this benchmark data set will encourage further research on person re-identification with clothes variation. Moreover, we also perform extensive analyses on this data set using several

show abstract

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

confidence: 99%

Section: Related Workmentioning

confidence: 99%

Section: Related Workmentioning

confidence: 99%

See 1 more Smart Citation

A benchmark for clothes variation in person re‐identification

Wang

Chen

et al. 2020

Int J Intell Syst

View full text Add to dashboard Cite

show abstract

“…Person re-identification aims is to search for a targeted person via surveillance videos at different locations and times. Due to factors like the limitations of technology, most of the current research on person re-identification assume that the target's clothes are unchanged (Huang et al, 2018;Jin et al, 2022;Li et al, 2018). Thus, it uses the color, texture, and other features of the clothes as discriminant conditions.…”

Section: Introductionmentioning

confidence: 99%

Dual-Branch Network Fused With Two-Level Attention Mechanism for Clothes-Changing Person Re-Identification

Jin

2023

International Journal of Web Services Research

View full text Add to dashboard Cite

Clothes-changing person re-identification is a hot topic in the current academic circles. Most of the current methods assume that the clothes of a person will not change in a short period of time, but they are not applicable when people change clothes. Based on this situation, this paper proposes a dual-branch network for clothes-changing person re-identification that integrates a two-level attention mechanism and captures and aggregates fine-grained person semantic information in channels and spaces through a two-level attention mechanism and suppresses the sensitivity of the network to clothing features by training the clothing classification branch. The method does not use auxiliary means such as human skeletons, and the complexity of the model is greatly reduced compared with most methods. This paper conducts experiments on the popular clothes-changing person re-identification dataset PRCC and a very large-scale cross-spatial-temporal dataset (LaST). The experimental results show that the method in this paper is more advanced than the existing methods.

show abstract

“…In recent years, plenty of efforts [3,19,21,36,47,50,53] have been made to handle the cloth-changing issue by learning discriminative cloth-agnostic identity representations. A small proportion of methods [3,47,50] attempt to decouple cloth-agnostic features directly from RGB images without multi-modal auxiliary information, which inevitably leads to the loss of crucial information in global features and results in a heavy reliance on the domain.…”

Section: Introductionmentioning

confidence: 99%

Exploring Shape Embedding for Cloth-Changing Person Re-Identification via 2D-3D Correspondences

Wang,

Yu,

Yan

et al. 2023

Proceedings of the 31st ACM International Conference on Multimedia

View full text Add to dashboard Cite

Cloth-Changing Person Re-Identification (CC-ReID) is a common and realistic problem since fashion constantly changes over time and people's aesthetic preferences are not set in stone. While most existing cloth-changing ReID methods focus on learning clothagnostic identity representations from coarse semantic cues (e.g. silhouettes and part segmentation maps), they neglect the continuous shape distributions at the pixel level. In this paper, we propose Continuous Surface Correspondence Learning (CSCL), a new shape embedding paradigm for cloth-changing ReID. CSCL establishes continuous correspondences between a 2D image plane and a canonical 3D body surface via pixel-to-vertex classification, which naturally aligns a person image to the surface of a 3D human model and simultaneously obtains pixel-wise surface embeddings. We further extract fine-grained shape features from the learned surface embeddings and then integrate them with global RGB features via a carefully designed cross-modality fusion module. The shape embedding paradigm based on 2D-3D correspondences remarkably enhances the model's global understanding of human body shape. To promote the study of ReID under clothing change, we construct 3D Dense Persons (DP3D), which is the first largescale cloth-changing ReID dataset that provides densely annotated 2D-3D correspondences and a precise 3D mesh for each person image, while containing diverse cloth-changing cases over all four seasons. Experiments on both cloth-changing and cloth-consistent ReID benchmarks validate the effectiveness of our method. Our project page is located at https://CSCL-CC.github.io.

show abstract

Cloth-Changing Person Re-identification from A Single Image with Gait Prediction and Regularization

Cited by 50 publications

References 50 publications

A benchmark for clothes variation in person re‐identification

A benchmark for clothes variation in person re‐identification

Dual-Branch Network Fused With Two-Level Attention Mechanism for Clothes-Changing Person Re-Identification

Exploring Shape Embedding for Cloth-Changing Person Re-Identification via 2D-3D Correspondences

Contact Info

Product

Resources

About