AD-Cluster: Augmented Discriminative Clustering for Domain Adaptive Person Re-Identification

Zhai, Yunpeng; Lu, Shijian; Ye, Qixiang; Shan, Xuebo; Chen, Jie; Ji, Rongrong; Tian, Yonghong

doi:10.1109/cvpr42600.2020.00904

Cited by 263 publications

(140 citation statements)

References 29 publications

Supporting

Mentioning

140

Contrasting

Order By: Relevance

“…We compare our proposed A 2 G framework with unsupervised, unsupervised domain adaptation, and attribute auxiliary weakly-supervised methods on four cross-dataset person re-ID tasks: Duke-to-Market, Market-to-Duke, MSMTto-Duke, and MSMT-to-Market. We compare three types of approaches, including unsupervised learning methods: PUAL [45], BUC [16], SSL [10], HCT [46], D-MMD [49], CSE [10], and TAUDL [47], transfer learning based methods: SPGAN [20], HHL [22], CFSM [48], ENC [15], UDATP [25], UCDA-CCE [50], PDA-Net [51], PCB-PAST [8], SSG [7], MMCL [24], DG-NET++ [52], B-SNR+GDS-H [53], DGNET [3], OG-Net [54], AE [17], and AD-Cluster [55], and attribute auxiliary weakly supervised method: TJ-AIDL [26].…”

Section: Comparison With the State-of-the-artsmentioning

confidence: 99%

Graph Neural Network Based Attribute Auxiliary Structured Grouping for Person Re-Identification

Tang

Gao²,

Chen

et al. 2024

IEEE Access

View full text Add to dashboard Cite

Recently, person re-identification (re-ID) with weakly labeled or unlabeled data has drawn much attention in open-set and cross-domain re-ID systems especially for the attribute auxiliary weakly supervised person re-ID. Although state-of-the-art clustering-based methods have achieved good performance, the pseudo labels generated through clustering are often low-quality and noisy. To address this problem, we propose a graph neural network based Attribute Auxiliary structured Grouping (A 2 G) to improve the confidence of the pseudo labels. Different from the existing clustering-based approaches that only utilize the similarity in feature space, we learn the feature representation from the similarities in both attribute space and feature space by graph learning on the pedestrian attribute graph. Specifically, we first utilize the pair-wise attribute similarity to represent the relation between two pedestrians to construct a pedestrian attribute graph. Furthermore, we aggregate the features from their neighborhood on a pedestrian attribute graph by the graph neural network, which would make the attribute similar pairs closer and simultaneously take the dissimilar pairs further in the feature space. Finally, to avoid the over-confidence of the hard pseudo labels, we regularize the learning of the embedding model with the smoothed pseudo label (SPL) in the optimization of the feature embedding network. We conduct extensive experiments on several person re-ID datasets to validate the efficacy of our proposed method. The results demonstrate that our technique is effective and promising for person re-ID tasks.INDEX TERMS Unsupervised person re-identification, attribute-auxiliary structured grouping, graph neural network.

show abstract

Section: Comparison With the State-of-the-artsmentioning

confidence: 99%

Graph Neural Network Based Attribute Auxiliary Structured Grouping for Person Re-Identification

Tang

Gao²,

Chen

et al. 2024

IEEE Access

View full text Add to dashboard Cite

show abstract

“…Yang et al [9] proposed self-similarity grouping from global to local methods to mine the potential similarity of unlabeled samples, automatically establish multiple clusters from different viewpoints, and label these independent clusters as pseudo-labels to supervise training. Zhai et al [10] used target domain sample generation to increase the clustering points, and hence the diversity of categories, and a feature encoder to minimize the distance between images within a class in the feature space to improve the accuracy of cross-domain person re-ID. Jin et al [11] tried to distinguish the distributions of positive and negative samples using a momentum update strategy during training.…”

Section: Introductionmentioning

confidence: 99%

Asymmetric Mutual Mean-Teaching for Unsupervised Domain Adaptive Person Re-Identification

Dong

Liu

2021

IEEE Access

View full text Add to dashboard Cite

Unsupervised domain adaptive person re-identification aims to solve the problem of poor performance caused by transferring an unlabeled target domain from the labeled source domain in the reidentification task. The clustering pseudo-labels method in unsupervised learning is widely used in unsupervised adaptive person re-identification tasks, and it maintains state-of-the-art performance. However, pseudo-labels obtained through clustering often have much noise, and the use of a single network model structure and a single clustering algorithm can easily cause model learning to stagnate, making the model not generalizable. To solve this problem, this paper proposes an asymmetric mutual mean-teaching method for unsupervised adaptive person re-identification. In terms of feature extraction, two asymmetric network models with different structures are used for mutual mean-teaching on the target domain, making the features extracted by the network more robust. In terms of feature clustering, two clustering methods are used for mutual teaching to dynamically update the centroid of clusters to improve the confidence of clustering pseudo-labels. Finally, the triplet loss is improved based on the updated cluster centroid to improve the clustering effect. The proposed method is used to perform a large number of verification experiments on three public datasets. The experimental results show that the proposed method has better accuracy than other unsupervised person re-identification based on clustering pseudo-labels.INDEX TERMS Deep learning, person re-identification, unsupervised domain adaptation, mutual meanteaching.

show abstract

“…Person re-identification(re-ID) aiming to establish person identity correspondence across different cameras in surveillance video, has attracted attention of many researchers and achieved impressive progress [1,2,3]in the past decade with the rapid evolution of deep learning. Although conventional supervised methods have achieved excellent performance, manually labeling datasets is a tedious and slow process.…”

Section: Introductionmentioning

confidence: 99%

“…The mainstream algorithms for UDA tasks lie in two aspects: image-style translated methods [4,5,1] and pseudolabels based methods [6,7,8,9,10,3]. The image-style * Liyan Zhang is the corresponding author.…”

Section: Introductionmentioning

confidence: 99%

“…As shown in Fig. (1), DBSCAN will split the target domain into inliers(high confidence samples) and outliers(low confidence samples) and we make full use of all the target domain samples including outliers instead of directly abandon these hard but valuable samples. We gradually add outliers to the training process in an adversarial strategy manner.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Rectifying Pseudo Label By Mutual Disagreement Learning For Unsupervised Domain Adaptation Person Re-Identification

Zhang

2021

2021 IEEE International Conference on Multimedia and Expo (ICME)

View full text Add to dashboard Cite

Unsupervised domain adaptation person re-identification(re-ID) aims to transfer knowledge from the labeled source domain to unlabeled target domain, which is still a challenging task due to the large domain discrepancy. The clusteringbased methods maintain advanced performance, generating pseudo-labels for unlabeled target domain images by clustering. However, not making full use of all valuable images and label noise derived from imperfect clustering results dramatically impact further performance improvement. To alleviate the above two problems, we propose a novel mutual disagreement learning(MDL) framework. We attach some of outliers(unclustered samples) with small loss to training process in an adversarial strategy manner. To rectify label noise, we train two networks and their momentum-based moving average models, making them teach each other and using prediction disagreement samples to update networks. Extensive experiments on four unsupervised domain tasks, Marketto-Duke, Duke-to-Market, Market-to-MSMT and Duke-to-MSMT, show that the advantages of our proposed MDL framework compared with other state-of-the-art methods.

show abstract

AD-Cluster: Augmented Discriminative Clustering for Domain Adaptive Person Re-Identification

Cited by 263 publications

References 29 publications

Graph Neural Network Based Attribute Auxiliary Structured Grouping for Person Re-Identification

Graph Neural Network Based Attribute Auxiliary Structured Grouping for Person Re-Identification

Asymmetric Mutual Mean-Teaching for Unsupervised Domain Adaptive Person Re-Identification

Rectifying Pseudo Label By Mutual Disagreement Learning For Unsupervised Domain Adaptation Person Re-Identification

Contact Info

Product

Resources

About