Reasoning and Tuning: Graph Attention Network for Occluded Person Re-Identification

Huang, Meiyan; Hou, Chunping; Yang, Qingyuan; Wang, Zhipeng

doi:10.1109/tip.2023.3247159

Cited by 28 publications

(3 citation statements)

References 62 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Where "-" represents that the authors of the original paper did not conduct relevant experiments on this dataset, but it can be seen that our method has better cross-domain generalization compared to some of the existing methods. The ACID [38] method adopts a global competitive strategy to deal with image noise, which artificially defines the noise distribution and reduces the effect of predefined noise by subsequent competitive learning, making it difficult to remove the real noise. In contrast, the feature self-enhancement module used in this paper uses the attention mechanism to automatically strengthen the important information and suppress the effect of noise, as can be seen from the experimental results, the method in this paper is more conducive to suppressing noise.…”

Section: Comparison Of Experimental Resultsmentioning

confidence: 99%

A method based on difference guide and feature self-enhancement for clothes-changing person re-identification

Ge,

Lu,

Xia

et al. 2024

Preprint

View full text Add to dashboard Cite

Due to the effect of clothing change on person re-identification models, some researchers have car-ried out in-depth studies on clothes-changing person re-identification(CC-ReID). However, there are some problem of the loss of edge identity information in the semantic guidance process in current methods. In this work, we propose a dual-stream network model, named GFSAnet, which consists of both global and face streams. This model is capable of retaining edge identity information while reinforcing the weight of fine-grained discriminative information. Firstly, in the global stream, we de-sign a difference guide model (DGM) and a feature self-augmentation model (FSAM). The differential features are learned through the difference guide module to preserve the edge identity information of the boundary between background and foreground, while the weights of the local information in the network are optimized through the feature self-augmentation module. Secondly, in the face stream, the multi-scale structure design of pyramid residual network is used to learn the facial features fusing coarse and fine granularity. Finally, the contributions of global and facial features are dynamically adjusted to work together in the inference by setting the hyperparameter α. Extensive experiments show that the method in this paper achieves better performance on the PRCC, Celeb-ReID and Celeb-Light datasets.

show abstract

Section: Comparison Of Experimental Resultsmentioning

confidence: 99%

A method based on difference guide and feature self-enhancement for clothes-changing person re-identification

Ge,

Lu,

Xia

et al. 2024

Preprint

View full text Add to dashboard Cite

show abstract

“…Result on Holistic Datasets. We also experiment our proposed method on holistic person Re-ID datasets, including Market-1501 and DukeMTMC-reID, and compare our method with state-of-the-art methods in three categories, i.e., holistic Re-ID methods (Zhu et al 2020;Sun et al 2018), partial Re-ID methods (Luo et al (Wang et al 2020;Chen et al 2021;He et al 2019;Li et al 2021;Wang et al 2022b;He et al 2021;Tan et al 2022;Wang et al 2022a;Jia et al 2023;Yan et al 2023;Zhao et al 2023) methods. The results are shown in Table 2.…”

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

confidence: 99%

Attention Disturbance and Dual-Path Constraint Network for Occluded Person Re-identification

Xia,

Tan,

Dai

et al. 2024

AAAI

View full text Add to dashboard Cite

Occluded person re-identification (Re-ID) aims to address the potential occlusion problem when matching occluded or holistic pedestrians from different camera views. Many methods use the background as artificial occlusion and rely on attention networks to exclude noisy interference. However, the significant discrepancy between simple background occlusion and realistic occlusion can negatively impact the generalization of the network. To address this issue, we propose a novel transformer-based Attention Disturbance and Dual-Path Constraint Network (ADP) to enhance the generalization of attention networks. Firstly, to imitate real-world obstacles, we introduce an Attention Disturbance Mask (ADM) module that generates an offensive noise, which can distract attention like a realistic occluder, as a more complex form of occlusion. Secondly, to fully exploit these complex occluded images, we develop a DualPath Constraint Module (DPC) that can obtain preferable supervision information from holistic images through dualpath interaction. With our proposed method, the network can effectively circumvent a wide variety of occlusions using the basic ViT baseline. Comprehensive experimental evaluations conducted on person re-ID benchmarks demonstrate the superiority of ADP over state-of-the-art 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. The mainstream methods [7,19,21,29,36,53] typically adopt human parsing models to obtain coarse semantic cues to guide the extraction of biometric features, such as shape features.…”

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

Reasoning and Tuning: Graph Attention Network for Occluded Person Re-Identification

Cited by 28 publications

References 62 publications

A method based on difference guide and feature self-enhancement for clothes-changing person re-identification

A method based on difference guide and feature self-enhancement for clothes-changing person re-identification

Attention Disturbance and Dual-Path Constraint Network for Occluded Person Re-identification

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

Contact Info

Product

Resources

About