PAFormer: Part Aware Transformer for Person Re-identification

TL;DR

PAFormer is a novel pose estimation based transformer model for person re-identification that improves part-to-part comparison accuracy without extra localization modules, leveraging learnable pose tokens and visibility prediction.

Contribution

It introduces pose tokens for precise part awareness and a visibility predictor trained with ground truth, enhancing partial ReID performance without additional localization modules.

Findings

Outperforms existing methods on ReID benchmarks.

Operates without extra body part localization modules.

Effectively handles occlusion with visibility prediction.

Abstract

Within the domain of person re-identification (ReID), partial ReID methods are considered mainstream, aiming to measure feature distances through comparisons of body parts between samples. However, in practice, previous methods often lack sufficient awareness of anatomical aspect of body parts, resulting in the failure to capture features of the same body parts across different samples. To address this issue, we introduce \textbf{Part Aware Transformer (PAFormer)}, a pose estimation based ReID model which can perform precise part-to-part comparison. In order to inject part awareness to pose tokens, we introduce learnable parameters called `pose token' which estimate the correlation between each body part and partial regions of the image. Notably, at inference phase, PAFormer operates without additional modules related to body part localization, which is commonly used in previous ReID methodologies leveraging pose estimation models. Additionally, leveraging the enhanced awareness of body parts, PAFormer suggests the use of a learning-based visibility predictor to estimate the degree of occlusion for each body part. Also, we introduce a teacher forcing technique using ground truth visibility scores which enables PAFormer to be trained only with visible parts. A set of extensive experiments show that our method outperforms existing approaches on well-known ReID benchmark datasets.