OOTDiffusion implements outfitting-fusion based latent diffusion for virtual try-on applications. The model takes a person image and a clothing garment image as input, then generates a new image with the garmentTry-on model synthesizes the garment onto the person while preserving pose and body structure. The system includes human parsing and pose estimation components, with separate models for half-body and full-body try-on scenarios. Checkpoints are released for VITON-HD and Dress Code datasets.