Unsupervised Long-Term Person Re-Identification with Clothes Change

TL;DR

This paper presents an unsupervised long-term person re-identification method that handles clothes changes without requiring identity labels, using a novel curriculum clustering approach to improve accuracy in real-world scenarios.

Contribution

We introduce a novel unsupervised long-term person re-identification framework with clothes change, utilizing curriculum clustering to adaptively improve clustering confidence and performance.

Findings

Outperforms state-of-the-art unsupervised re-id methods

Closely matches supervised re-id models in accuracy

Effective on three long-term re-id datasets

Abstract

We investigate unsupervised person re-identification (Re-ID) with clothes change, a new challenging problem with more practical usability and scalability to real-world deployment. Most existing re-id methods artificially assume the clothes of every single person to be stationary across space and time. This condition is mostly valid for short-term re-id scenarios since an average person would often change the clothes even within a single day. To alleviate this assumption, several recent works have introduced the clothes change facet to re-id, with a focus on supervised learning person identity discriminative representation with invariance to clothes changes. Taking a step further towards this long-term re-id direction, we further eliminate the requirement of person identity labels, as they are significantly more expensive and more tedious to annotate in comparison to short-term person re-id datasets. Compared to conventional unsupervised short-term re-id, this new problem is drastically more challenging as different people may have similar clothes whilst the same person can wear multiple suites of clothes over different locations and times with very distinct appearance. To overcome such obstacles, we introduce a novel Curriculum Person Clustering (CPC) method that can adaptively regulate the unsupervised clustering criterion according to the clustering confidence. Experiments on three long-term person re-id datasets show that our CPC outperforms SOTA unsupervised re-id methods and even closely matches the supervised re-id models.