Adversarial Attribute-Image Person Re-identification

TL;DR

This paper introduces a novel joint space learning approach for cross-modality person re-identification, matching images to attribute descriptions using attribute-guided attention and semantic adversarial strategies, outperforming existing methods.

Contribution

It proposes a new joint space learning framework with attribute-guided attention and semantic adversarial training for cross-modality person Re-ID, addressing a rarely studied problem.

Findings

Outperforms existing methods on three attribute datasets.

Effective in learning semantically correlated features across modalities.

Demonstrates robustness in attribute-image matching for surveillance applications.

Abstract

While attributes have been widely used for person re-identification (Re-ID) which aims at matching the same person images across disjoint camera views, they are used either as extra features or for performing multi-task learning to assist the image-image matching task. However, how to find a set of person images according to a given attribute description, which is very practical in many surveillance applications, remains a rarely investigated cross-modality matching problem in person Re-ID. In this work, we present this challenge and formulate this task as a joint space learning problem. By imposing an attribute-guided attention mechanism for images and a semantic consistent adversary strategy for attributes, each modality, i.e., images and attributes, successfully learns semantically correlated concepts under the guidance of the other. We conducted extensive experiments on three attribute datasets and demonstrated that the proposed joint space learning method is so far the most effective method for the attribute-image cross-modality person Re-ID problem.