Graph-Based Self-Learning for Robust Person Re-Identification

Yuqiao Xian, Jinrui Yang, Fufu Yu, Jun Zhang, Xing Sun; Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision (WACV), 2023, pp. 4789-4798


Existing deep learning approaches for person re-identification (Re-ID) mostly rely on large-scale and well-annotated training data. However, human-annotated labels are prone to label noise in real-world applications. Previous person Re-ID works mainly focus on random label noise, which doesn't properly reflect the characteristic of label noise in practical human-annotated process. In this work, we find the visual ambiguity noise is more common and reasonable noise assumption in annotation of person Re-ID. To handle the kind of noise, we propose a simple and effective robust person Re-ID framework, namely Graph-Based Self-Learning (GBSL), to iteratively learn discriminative representation and rectify noisy labels with limited annotated samples for each identity. Meanwhile, considering the practical annotation process in person Re-ID, we further extend the visual ambiguity noise assumption and propose a type of more practical label noise in person Re-ID, namely the tracklet-level label noise (TLN). Without modifying network architecture or loss function, our approach significantly improves the robustness against label noise of the Re-ID system. Our model obtains competitive performance with training data corrupted by various types of label noise and outperforms the existing methods for robust Re-ID on public benchmarks.

Related Material

[pdf] [supp]
@InProceedings{Xian_2023_WACV, author = {Xian, Yuqiao and Yang, Jinrui and Yu, Fufu and Zhang, Jun and Sun, Xing}, title = {Graph-Based Self-Learning for Robust Person Re-Identification}, booktitle = {Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision (WACV)}, month = {January}, year = {2023}, pages = {4789-4798} }