Unsupervised Domain Adaptive Re-Identification for reid, with clustering and noisy label techniques.
Implementation of the paper Unsupervised Domain Adaptive Re-Identification: Theory and Practice, and [Probabilistic End-to-end Noise Correction for Learning with Noisy Labels][http://openaccess.thecvf.com/content_CVPR_2019/papers/Yi_Probabilistic_End-To-End_Noise_Correction_for_Learning_With_Noisy_Labels_CVPR_2019_paper.pdf ] The selftraining scheme proposed in the paper is simple yet effective.
Our code is based on open-reid. [SSG][https://github.com/OasisYang/SSG], and [PENCIL][https://github.com/yikun2019/PENCIL/blob/master/PENCIL.py]