DreamID-V:Bridging the Image-to-Video Gap for High-Fidelity Face Swapping via Diffusion Transformer Paper • 2601.01425 • Published 8 days ago • 48