Refining Pseudo Labels with Clustering Consensus over Generations for Unsupervised Object Re-identification

Zhang, Xiao; Ge, Yixiao; Qiao, Yu; Li, Hongsheng

doi:10.1109/cvpr46437.2021.00344

Cited by 101 publications

(36 citation statements)

References 31 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…To tackle the challenge of pseudo label noise an approach was presented in [255]. They have refined the pseudo labels based on clustering consensus.…”

Section: Cnn-based Approachesmentioning

confidence: 99%

Person Re-identification: A Retrospective on Domain Specific Open Challenges and Future Trends

Zahra¹,

Perwaiz²,

Shahzad³

et al. 2022

Preprint

View full text Add to dashboard Cite

Person re-identification (Re-ID) is one of the primary components of an automated visual surveillance system. It aims to automatically identify/search persons in a multi-camera network having non-overlapping field-of-views. Owing to its potential in various applications and research significance, a plethora of deep learning based re-Id approaches have been proposed in the recent years. However, there exist several vision related challenges, e.g., occlusion, pose scale & viewpoint variance, background clutter, person misalignment and cross-domain generalization across camera modalities, which makes the problem of re-Id still far from being solved. Majority of the proposed approaches directly or indirectly aim to solve one or multiple of these existing challenges. In this context, a comprehensive review of current re-ID approaches in solving theses challenges is needed to analyze and focus on particular aspects for further advancements. At present, such a focused review does not exist and henceforth in this paper, we have presented a systematic challenge-specific literature survey of 230+ papers between the years of 2015-21. For the first time a survey of this type have been presented where the person re-Id approaches are reviewed in such solution-oriented perspective. Moreover, we have presented several diversified prominent developing trends in the respective research domain which will provide a visionary perspective regarding ongoing person re-Id research and eventually help to develop practical real world solutions. Scope/Objective of the ReviewIn this paper, we have targeted the most popular challenges in person re-id to perform systematic challenge-wise review of the published approaches. In this context, we have collected papers from top conferences and journals for the years from 2015 to 2021. The progress in papers addressing each challenge and its influence on published results is

show abstract

“…To tackle the challenge of pseudo label noise an approach was presented in [255]. They have refined the pseudo labels based on clustering consensus.…”

Section: Cnn-based Approachesmentioning

confidence: 99%

Person Re-identification: A Retrospective on Domain Specific Open Challenges and Future Trends

Zahra¹,

Perwaiz²,

Shahzad³

et al. 2022

Preprint

View full text Add to dashboard Cite

show abstract

“…They also minimize the loss function considering cluster centroids and feature samples stored in feature memory. RLCC [9] refines clusters by a consensus among iterations. Pseudo-labels on a certain iteration are created by considering the ones generated on a previous iteration, keeping the training stable.…”

Section: B Person Re-identificationmentioning

confidence: 99%

“…This clearly shows that Person ReID is a more demanding task than ImageNet classification, and general state-of-the-art self-supervised learning methods are not suitable for the task. These methods result in much less accurate models when compared to prior unsupervised learning methods tailored specifically to ReID, even with ImageNet weight initialization [9], [10].…”

Section: Introductionmentioning

confidence: 99%

Reasoning for Complex Data through Ensemble-based Self-Supervised Learning

Bertocco¹,

Theophilo²,

Andaló³

et al. 2022

Preprint

View full text Add to dashboard Cite

“…Various CL-based methods [6], [10]- [12], [18], [31], [35] have been developed for unsupervised Re-ID. Moreover, the issue of pseudo label noise occurred in unsupervised Re-ID has also been noticed recently and various methods [14], [16], [18]- [20] have been proposed to address it. We will introduce these related work and state the difference of our work in Section II-B and II-D.…”

Section: Introductionmentioning

confidence: 99%

“…It implies that the generated pseudo labels might be out of date. To deal with the inevitable label noise, efforts have been made in label refinement [14]- [16], hybrid contrastive learning [10], [12], [17] that combines cluster-and instance-level contrasts together, online pseudo label generation [18], and other techniques [19], [20].…”

Section: Introductionmentioning

confidence: 99%

Offline-Online Associated Camera-Aware Proxies for Unsupervised Person Re-identification

Wang¹,

Li²,

Lai³

et al. 2022

Preprint

View full text Add to dashboard Cite

Recently, unsupervised person re-identification (Re-ID) has received increasing research attention due to its potential for label-free applications. A promising way to address unsupervised Re-ID is clustering-based, which generates pseudo labels by clustering and uses the pseudo labels to train a Re-ID model iteratively. However, most clustering-based methods take each cluster as a pseudo identity class, neglecting the intra-cluster variance mainly caused by the change of cameras. To address this issue, we propose to split each single cluster into multiple proxies according to camera views. The camera-aware proxies explicitly capture local structures within clusters, by which the intra-ID variance and inter-ID similarity can be better tackled. Assisted with the camera-aware proxies, we design two proxylevel contrastive learning losses that are, respectively, based on offline and online association results. The offline association directly associates proxies according to the clustering and splitting results, while the online strategy dynamically associates proxies in terms of up-to-date features to reduce the noise caused by the delayed update of pseudo labels. The combination of two losses enable us to train a desirable Re-ID model. Extensive experiments on three person Re-ID datasets and one vehicle Re-ID dataset show that our proposed approach demonstrates competitive performance with state-of-the-art methods. Code will be available at: https://github.com/Terminator8758/O2CAP.

show abstract

Refining Pseudo Labels with Clustering Consensus over Generations for Unsupervised Object Re-identification

Cited by 101 publications

References 31 publications

Person Re-identification: A Retrospective on Domain Specific Open Challenges and Future Trends

Person Re-identification: A Retrospective on Domain Specific Open Challenges and Future Trends

Reasoning for Complex Data through Ensemble-based Self-Supervised Learning

Offline-Online Associated Camera-Aware Proxies for Unsupervised Person Re-identification

Contact Info

Product

Resources

About