DreamID-V:Bridging the Image-to-Video Gap for High-Fidelity Face Swapping via Diffusion Transformer
IntermediateXu Guo, Fulong Ye et al.Jan 4arXiv
DreamID-V is a new AI method that swaps faces in videos while keeping the body movements, expressions, lighting, and background steady and natural.
#video face swapping#image face swapping#diffusion transformer