FW-VTON: Flattening-and-Warping for Person-to-Person Virtual Try-On

Traditional virtual try-on mainly addresses garment-to-person scenarios that require flat garment representations. We study person-to-person try-on, which takes as input a target person and a source image of another individual wearing the desired garment, and aims to synthesize the target dressed in that garment. The key challenge is robustly extracting the garment despite pose changes and occlusions and realistically adapting it to the target person. We present FW-VTON, a three-stage framework: (1) garment flattening to reconstruct a canonical, pose-agnostic garment from the source; (2) garment warping to align the flattened garment with the target pose; and (3) seamless integration onto the target person. We also introduce a benchmark dataset tailored to this task. Extensive experiments demonstrate that FW-VTON achieves state-of-the-art performance in both quantitative metrics and visual quality, and also on the garment extraction subtask.