Deviation based clustering for unsupervised person re-identification

Ning, Munan; Zeng, Kaiwei; Guo, Yang; Wang, Yaohua

doi:10.1016/j.patrec.2020.04.039

Cited by 26 publications

(45 citation statements)

References 4 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…We also test our method in unsupervised scenario, i.e., training the ReID model only with GLO. Our method achieves Rank-1 accuracy of 77.4% on Market-1501, significantly outperforming recent BUC [25] and DBC [6] by 11.2% and 8.2%, respectively. It is worth noting that, our unsupervised training also outperforms several domain adaptive methods that use extra source domain for training, such as PAUL [45] and DA_2S [14].…”

Section: Figure 1: Illustration Of Proposed Coupling Optimization Metmentioning

confidence: 84%

See 1 more Smart Citation

Domain Adaptive Person Re-Identification via Coupling Optimization

Liu

Zhang

2020

Proceedings of the 28th ACM International Conference on Multimedia

View full text Add to dashboard Cite

Domain adaptive person Re-Identification (ReID) is challenging owing to the domain gap and shortage of annotations on target scenarios. To handle those two challenges, this paper proposes a coupling optimization method including the Domain-Invariant Mapping (DIM) method and the Global-Local distance Optimization (GLO), respectively. Different from previous methods that transfer knowledge in two stages, the DIM achieves a more efficient one-stage knowledge transfer by mapping images in labeled and unlabeled datasets to a shared feature space. GLO is designed to train the ReID model with unsupervised setting on the target domain. Instead of relying on existing optimization strategies designed for supervised training, GLO involves more images in distance optimization, and achieves better robustness to noisy label prediction. GLO also integrates distance optimizations in both the global dataset and local training batch, thus exhibits better training efficiency. Extensive experiments on three large-scale datasets, i.e., Market-1501, DukeMTMC-reID, and MSMT17, show that our coupling optimization outperforms state-of-the-art methods by a large margin. Our method also works well in unsupervised training, and even outperforms several recent domain adaptive methods.

show abstract

Section: Figure 1: Illustration Of Proposed Coupling Optimization Metmentioning

confidence: 84%

“…Comparisons are performed on Market-1501 and DukeMTMC-reID. Compared methods include LOMO [24], BOW [54], DBC [6], and BUC [25]. It can be observed from Table 3 that proposed GLO achieves the best performance on both datasets and outperforms state-of-the-art methods by a clear margin.…”

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

confidence: 96%

Domain Adaptive Person Re-Identification via Coupling Optimization

Liu

Zhang

2020

Proceedings of the 28th ACM International Conference on Multimedia

View full text Add to dashboard Cite

show abstract

“…As the popular technology in unlabeled pedestrian image re-id, many researches [3], [17], [18], [20], [30], [40], [46], [48] [30] formulated an unsupervised multi-label classification for per-…”

Section: ) Comparison With Unsupervised Clustering Methodsmentioning

confidence: 99%

“…This paper conducts comparison between unsupervised domain adaptation (UDA) [5], [7], [8], [16], [23] and clustering methods [3], [17], [18], [20], [30], [40], [46], [48]. The re-id results of these methods are reported in it proposed a multi-task dictionary learning method to learn a dataset-shared but target-data-biased person representation, which outperforms the former state-of-the-arts; Li et al [16] exploited useful knowledge of pre-existing labeled data from…”

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

confidence: 99%

Unsupervised Person Re-Identification via Multi-Order Cross-View Graph Adversarial Network

Lai

2021

IEEE Access

View full text Add to dashboard Cite

“…Recent methods which follow this approach are [6,7,33,35,36]. Completely unsupervised methods, including [5,18] do not include supervision at any stage of training. Some of these works propose generating pseudo-labels [7,18,35] or pseudo-positive samples [33] that can be utilized as previously within classification or metric losses, respectively.…”

Section: Supervised Methodsmentioning

confidence: 99%

Unsupervised Domain Adaptation for Person Re-Identification with Few and Unlabeled Target Data

Galanakis

Zabulis

Argyros

2020

Advances in Visual Computing

View full text Add to dashboard Cite

Existing, fully supervised methods for person re-identification (ReID) require annotated data acquired in the target domain in which the method is expected to operate. This includes the IDs as well as images of persons in that domain. This is an obstacle in the deployment of ReID methods in novel settings. For solving this problem, semi-supervised or even unsupervised ReID methods have been proposed. Still, due to their assumptions and operational requirements, such methods are not easily deployable and/or prove less performant to novel domains/settings, especially those related to small person galleries. In this paper, we propose a novel approach for person ReID that alleviates these problems. This is achieved by proposing a completely unsupervised method for fine tuning the ReID performance of models learned in prior, auxiliary domains, to new, completely different ones. The proposed model adaptation is achieved based on only few and unlabeled target persons' data. Extensive experiments investigate several aspects of the proposed method in an ablative study. Moreover, we show that the proposed method is able to improve considerably the performance of state-of-the-art ReID methods in state-of-the-art datasets.

show abstract

Deviation based clustering for unsupervised person re-identification

Cited by 26 publications

References 4 publications

Domain Adaptive Person Re-Identification via Coupling Optimization

Domain Adaptive Person Re-Identification via Coupling Optimization

Unsupervised Person Re-Identification via Multi-Order Cross-View Graph Adversarial Network

Unsupervised Domain Adaptation for Person Re-Identification with Few and Unlabeled Target Data

Contact Info

Product

Resources

About