Chat-Driven Text Generation and Interaction for Person Retrieval
This addresses scalability and practical deployment challenges in surveillance applications by reducing annotation costs for text-based person retrieval.
The paper tackles the labor-intensive annotation bottleneck in text-based person search by introducing a unified annotation-free framework with two modules: Multi-Turn Text Generation for creating pseudo-labels via simulated dialogues, and Multi-Turn Text Interaction for refining vague user queries at inference. The method achieves competitive or superior retrieval accuracy while eliminating manual captions.
Text-based person search (TBPS) enables the retrieval of person images from large-scale databases using natural language descriptions, offering critical value in surveillance applications. However, a major challenge lies in the labor-intensive process of obtaining high-quality textual annotations, which limits scalability and practical deployment. To address this, we introduce two complementary modules: Multi-Turn Text Generation (MTG) and Multi-Turn Text Interaction (MTI). MTG generates rich pseudo-labels through simulated dialogues with MLLMs, producing fine-grained and diverse visual descriptions without manual supervision. MTI refines user queries at inference time through dynamic, dialogue-based reasoning, enabling the system to interpret and resolve vague, incomplete, or ambiguous descriptions - characteristics often seen in real-world search scenarios. Together, MTG and MTI form a unified and annotation-free framework that significantly improves retrieval accuracy, robustness, and usability. Extensive evaluations demonstrate that our method achieves competitive or superior results while eliminating the need for manual captions, paving the way for scalable and practical deployment of TBPS systems.