Asymmetric network pseudo labels mutual refinement for unsupervised domain adaptation person re-identification

被引：1
作者：
Yun X. [1 ]
Chen J. [1 ]
Zhang X. [1 ]
Dong K. [1 ]
Li S. [1 ,2 ]
Sun Y. [1 ]
机构：
[1] School of Information and Control Engineering, China University of Mining and Technology, Xuzhou
[2] Xi’an Key Laboratory of Heterogeneous Network Convergence Communication, Xi’an
来源：
Multimedia Tools and Applications | 2024年 / 83卷 / 40期
关键词：
Asymmetric network; Feature space loss; Pseudo labels mutual refinement; Unsupervised domain adaptation;
D O I：
10.1007/s11042-024-18912-7
中图分类号：
学科分类号：
摘要：
In the task of unsupervised domain adaptation person re-identification, the traditional symmetric dual-branch network only generates one single feature, which ignores the difference and complementarity of the network and is prone to produce wrong cluster results. To solve this problem, the asymmetric network pseudo labels mutual refinement (ANPR) algorithm is proposed to design an asymmetric network (ASNet), containing diverse features for mutual supervision. One branch of ASNet employs convolutional operations with constrained receptive fields to extract local information, and the other branch uses the contextual transformer block to capture global features. Secondly, this paper constructs the pseudo labels mutual refinement (PMR) module, which generates two sets of clustering results using global and local potential features of unlabeled samples. PMR suppresses the inconsistent clustering results and retains the consistent ones, which gradually improves the quality of pseudo labels for dual-brach mutual supervision. In addition, the feature space loss (FSL) function is designed to focus on the relative distance of samples in the feature layer without relying on the pseudo labels in the task layer, which effectively avoids the influence of noisy pseudo labels in the optimization process. The proposed method is evaluated on three popular datasets, and extensive experimental results demonstrate its effectiveness. © The Author(s), under exclusive licence to Springer Science+Business Media, LLC, part of Springer Nature 2024.
引用
收藏
页码：88091 / 88111
页数：20
相关论文
共 51 条
  • [1] Fang W., Yi W., Pang L., Sheng V.S., Study of cross-domain person re-identification based on dcgan, Multimed Tools Appl, 81, 25, pp. 36551-36565, (2022)
  • [2] Han J., Li Y.-L., Wang S., Delving into probabilistic uncertainty for unsupervised domain adaptive person re-identification, Proc AAAI Conf Artif Intell, 36, pp. 790-798, (2022)
  • [3] Chen S., Qiu L., Tian Z., Yan Y., Wang D.-H., Zhu S., Mtnet: mutual tri-training network for unsupervised domain adaptation on person re-identification, J Vis Commun Image Represent, 90, (2023)
  • [4] Rani J.S.J., Augasta M.G., Poolnet deep feature based person re-identification, Multimed Tools Appl, 82, 16, pp. 24967-24989, (2023)
  • [5] He L., Liang J., Li H., Sun Z., Deep spatial feature reconstruction for partial person re-identification: Alignment-free approach, In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pp. 7073-7082, (2018)
  • [6] Wang G., Yuan Y., Chen X., Li J., Zhou X., Learning discriminative features with multiple granularities for person re-identification, In: Proceedings of the 26Th ACM International Conference on Multimedia., pp. 274-282, (2018)
  • [7] Zhang X., Jing X.-Y., Zhu X., Ma F., Semi-supervised person re-identification by similarity-embedded cycle gans, Neural Comput Appl, 32, pp. 14143-14152, (2020)
  • [8] Chong Y., Peng C., Zhang J., Pan S., Style transfer for unsupervised domain-adaptive person re-identification, Neurocomputing, 422, pp. 314-321, (2021)
  • [9] Isola P., Zhu J.-Y., Zhou T., Efros A.A., Image-to-image translation with conditional adversarial networks, In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition., pp. 1125-1134, (2017)
  • [10] Long M., Cao Y., Wang J., Jordan M., Learning transferable features with deep adaptation networks, In: International Conference on Machine Learning, pp. 97-105, (2015)
123456