DiffV2S: Diffusion-based Video-to-Speech Synthesis with Vision-guided Speaker Embedding.

Jeongsoo Choi, Joanna Hong, Yong Man Ro

Browse the full ICCV paper archive.