Generalizable Person Re-identification Without Demographics

YiFan Zhang, Feng Li, Zhang Zhang, Baosheng Yu, Liang Wang, Dacheng Tao, Tieniu Tan

Keywords: Generalizable Person Re-Identification, Distributionally robust optimization, Change-of-measure technique

Abstract: Domain generalizable person re-identification (DG-ReID) aims to learn a ready-to-use domain-agnostic model directly for cross-dataset/domain evaluation, while current methods mainly explore the demographic information such as domain and/or camera labels for domain-invariant representation learning. However, the above-mentioned demographic information is not always accessible in practice due to privacy and security issues. In this paper, we consider the problem of person re-identification in a more general setting, \ie domain generalizable person re-identification without demographics (\textbf{DGWD-ReID}). To address the underlying uncertainty of domain distribution, we introduce distributionally robust optimization (DRO) to learn robust person re-identification models that perform well on all possible data distributions within the uncertainty set without demographics. However, directly applying the popular Kullback-Leibler divergence constrained DRO (or KL-DRO) fails to generalize well under the distribution shifts in real-world scenarios, since the convex condition may not hold for overparameterized neural networks. Inspired by this, we analyze and reformulate the popular KL-DRO by applying the change-of-measure technique, and then propose a simple yet efficient approach, \textbf{Unit-DRO}, which minimizes the loss over a new dataset with hard samples upweighted and other samples downweighted. We perform extensive experiments on both domain generalizable and cross-domain person re-identification tasks, and the empirical results on several large-scale benchmarks show that \iw~achieves superior performance compared to all baselines without using demographics.

Anonymous Url: I certify that there is no URL (e.g., github page) that could be used to find authors’ identity.

No Acknowledgement Section: I certify that there is no acknowledgement section in this submission for double blind review.

Code Of Ethics: I acknowledge that I and all co-authors of this work have read and commit to adhering to the ICLR Code of Ethics

Submission Guidelines: Yes

Please Choose The Closest Area That Your Submission Falls Into: Applications (eg, speech processing, computer vision, NLP)