D 3 F: Diffusion-Driven Dual-Stream Framework for Occluded Person Re-Identification
A novel end-to-end Diffusion-Driven Dual-stream Framework (D 3 F), which seamlessly integrates generative structural priors from Diffusion Transformers (DiT) into vision-language ReID, achieving state-of-the-art (SOTA) performance on both occluded and holistic ReID benchmark datasets.
Xiaohao Xie, Weihao Meng, Wen-Hua Jiao
· 0 citations