iTryOn: Mastering Interactive Video Virtual Try-On with Spatial-Semantic Guidance

Researchers introduced Interactive VVT and iTryOn, a video diffusion Transformer for interactive virtual try-on in videos.

The method adds multi-level interaction injection, 3D hand priors, and action captions to better model active clothing changes.

Action-aware positional embeddings improve spatial and temporal control, helping preserve consistency during garment deformation.

The result extends virtual try-on from passive image editing to realistic interactive video settings.