DRFormer: A Dual-Regularized Bidirectional Transformer for Person Re-identification

TL;DR

DRFormer integrates local texture and global semantic features using a dual-regularized bidirectional transformer, effectively improving person re-identification performance by leveraging the complementary strengths of vision foundation and vision-language models.

Contribution

The paper introduces DRFormer, a novel framework that synergizes vision foundation and vision-language models through dual-regularization for enhanced person re-identification.

Findings

Achieves competitive results on five benchmarks.

Effectively balances local and global feature contributions.

Demonstrates the benefit of integrating different model paradigms.

Abstract

Both fine-grained discriminative details and global semantic features can contribute to solving person re-identification challenges, such as occlusion and pose variations. Vision foundation models (\textit{e.g.}, DINO) excel at mining local textures, and vision-language models (\textit{e.g.}, CLIP) capture strong global semantic difference. Existing methods predominantly rely on a single paradigm, neglecting the potential benefits of their integration. In this paper, we analyze the complementary roles of these two architectures and propose a framework to synergize their strengths by a \textbf{D}ual-\textbf{R}egularized Bidirectional \textbf{Transformer} (\textbf{DRFormer}). The dual-regularization mechanism ensures diverse feature extraction and achieves a better balance in the contributions of the two models. Extensive experiments on five benchmarks show that our method effectively harmonizes local and global representations, achieving competitive performance against state-of-the-art methods.