Text Based Person Retrieval With Noisy

Text-based Person Retrieval with Noisy Correspondence is a benchmark task in the field of computer vision, aiming to retrieve specific individuals through textual descriptions in the presence of noisy correspondence. The core objective of this task is to improve the robustness and accuracy of models in imperfect data environments, thereby enhancing performance in practical application scenarios such as surveillance video analysis and social media image search. All recorded results are tested under a 20% noise rate.