tryoffdiff and try-off-anyone
These are competitors: both reconstruct and generate garment imagery from dressed persons, but TryOffDiff uses diffusion-based high-fidelity reconstruction while TryOffAnyone uses tiled cloth generation, offering different technical approaches to the same virtual try-off task.
About tryoffdiff
rizavelioglu/tryoffdiff
[CVPR'25-Demo] Official repository of "TryOffDiff: Virtual-Try-Off via High-Fidelity Garment Reconstruction using Diffusion Models".
Leverages Stable Diffusion v1.4 with SigLIP image encoding and multi-GPU training via HuggingFace Accelerate to reconstruct garments at high fidelity, supporting both single and multi-garment virtual try-off scenarios. The codebase provides modular training, inference, and evaluation pipelines compatible with VITON-HD and Dress Code datasets, with metrics from IQA-PyTorch, clean-fid, and DISTS for benchmarking reconstruction quality.
About try-off-anyone
ixarchakos/try-off-anyone
Official repository of "TryOffAnyone: Tiled Cloth Generation from a Dressed Person"
Scores updated daily from GitHub, PyPI, and npm data. How scores work