Take More Positives (TMP) is a contrastive learning framework for unsupervised person re-identification (re-ID) On the Market-1501 benchmark, TMP achieves 88.3% Rank-1 accuracy and 70.4% mean average precision . The framework of TMP outperforms the state-of-the-art methods for re-ID . TMP generates pseudo-labels by clustering samples, and updates itself with such pseudo labels and the proposed contrastive loss . The code will be made publicly available .

Author(s) : Xuanyu He, Wei Zhang, Ran Song, Xiangyuan Lan

Links : PDF - Abstract

Code :

Keywords : tmp - contrastive - framework - pseudo - identification -

Leave a Reply

Your email address will not be published. Required fields are marked *