End-to-End Context-Aided Unicity Matching for Person Re-identification

TL;DR

This paper introduces an end-to-end context-aware unicity matching approach for person re-identification that leverages global and contextual information to improve matching accuracy and efficiency across multiple datasets.

Contribution

It proposes a novel unicity matching architecture that refines person matching relations using graph neural networks and bipartite graph matching, applicable in both one-shot and multi-shot scenarios.

Findings

Outperforms existing methods on five public benchmarks.

Achieves high accuracy with a fast unicity matching variant.

Demonstrates robustness in both one-shot and multi-shot settings.

Abstract

Most existing person re-identification methods compute the matching relations between person images across camera views based on the ranking of the pairwise similarities. This matching strategy with the lack of the global viewpoint and the context's consideration inevitably leads to ambiguous matching results and sub-optimal performance. Based on a natural assumption that images belonging to the same person identity should not match with images belonging to multiple different person identities across views, called the unicity of person matching on the identity level, we propose an end-to-end person unicity matching architecture for learning and refining the person matching relations. First, we adopt the image samples' contextual information in feature space to generate the initial soft matching results by using graph neural networks. Secondly, we utilize the samples' global context relationship to refine the soft matching results and reach the matching unicity through bipartite graph matching. Given full consideration to real-world person re-identification applications, we achieve the unicity matching in both one-shot and multi-shot settings of person re-identification and further develop a fast version of the unicity matching without losing the performance. The proposed method is evaluated on five public benchmarks, including four multi-shot datasets MSMT17, DukeMTMC, Market1501, CUHK03, and a one-shot dataset VIPeR. Experimental results show the superiority of the proposed method on performance and efficiency.