TY  - RPRT
TI  - FLOAT: Generative Motion Latent Flow Matching for Audio-driven Talking Portrait
AU  - Taekyung Ki
AU  - Dongchan Min
AU  - Gyeongsu Chae
PY  - 2025
UR  - https://arxiv.org/abs/2412.01064
ID  - 2412.01064
ER  -