Multi-Scale Body-Part Mask Guided Attention for Person Re-identification

TL;DR

This paper introduces a multi-scale body-part mask guided attention network for person re-identification, effectively handling pose variation, misalignment, and background clutter, leading to state-of-the-art results.

Contribution

It proposes a novel attention network that jointly learns global and local features guided by body-part masks for improved person re-identification.

Findings

Achieves 95.0% rank-1 accuracy on Market1501

Attains 89.5% rank-1 accuracy on DukeMTMC-reID

Outperforms existing state-of-the-art methods

Abstract

Person re-identification becomes a more and more important task due to its wide applications. In practice, person re-identification still remains challenging due to the variation of person pose, different lighting, occlusion, misalignment, background clutter, etc. In this paper, we propose a multi-scale body-part mask guided attention network (MMGA), which jointly learns whole-body and part body attention to help extract global and local features simultaneously. In MMGA, body-part masks are used to guide the training of corresponding attention. Experiments show that our proposed method can reduce the negative influence of variation of person pose, misalignment and background clutter. Our method achieves rank-1/mAP of 95.0%/87.2% on the Market1501 dataset, 89.5%/78.1% on the DukeMTMC-reID dataset, outperforming current state-of-the-art methods.