Evolution of ReID: From Early Methods to LLM Integration

2025-06-16

AI Summary

This paper addresses the limited robustness of person re-identification (ReID) under illumination, pose, and viewpoint variations by proposing the first large language model (LLM)-enhanced cross-modal ReID framework. Methodologically, it innovatively leverages GPT-4o to generate identity-specific dynamic prompts, enabling fine-grained, identity-level alignment between images and textual descriptions; it integrates a visual encoder (ResNet/ViT), a cross-modal alignment module, and prompt engineering to deeply infuse natural language semantics into visual matching. Key contributions include: (1) the first systematic survey of LLM-driven cross-modal ReID paradigms; (2) the construction and open-sourcing of a large-scale ReID text description dataset generated by GPT-4o; and (3) empirical validation demonstrating substantial improvements in matching accuracy—particularly in challenging scenarios involving motion blur and occlusion.

Technology Category

• Machine Learning: Large Multimodal Models (LMMs)
• Computer Vision: Language and Vision
• Natural Language Processing: Language Grounding & Multi-modal NLP

Application Category

• User Modeling, Personalization and Recommendation: Large Language Models (LLM) for user modeling and recommendation
• Search and Retrieval-Augmented AI: Large language models for search
• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions

Abstract

Person re-identification (ReID) has evolved from handcrafted feature-based methods to deep learning approaches and, more recently, to models incorporating large language models (LLMs). Early methods struggled with variations in lighting, pose, and viewpoint, but deep learning addressed these issues by learning robust visual features. Building on this, LLMs now enable ReID systems to integrate semantic and contextual information through natural language. This survey traces that full evolution and offers one of the first comprehensive reviews of ReID approaches that leverage LLMs, where textual descriptions are used as privileged information to improve visual matching. A key contribution is the use of dynamic, identity-specific prompts generated by GPT-4o, which enhance the alignment between images and text in vision-language ReID systems. Experimental results show that these descriptions improve accuracy, especially in complex or ambiguous cases. To support further research, we release a large set of GPT-4o-generated descriptions for standard ReID datasets. By bridging computer vision and natural language processing, this survey offers a unified perspective on the field's development and outlines key future directions such as better prompt design, cross-modal transfer learning, and real-world adaptability.

Problem

• Evolution of ReID from handcrafted to LLM-integrated methods
• Addressing lighting, pose, viewpoint variations in person re-identification
• Improving visual matching using LLM-generated textual descriptions

Innovation

• Integrate LLMs for semantic and contextual information
• Use GPT-4o for dynamic identity-specific prompts
• Generate textual descriptions to enhance visual matching