Official SemTalk, EchoMask, GlobalDiff, and StreamTalk papers, models, and data for co-speech gesture generation.
-
SemTalk: Holistic Co-speech Motion Generation with Frame-level Semantic Emphasis
Paper • 2412.16563 • Published -
X-Zhang/SemTalk
Updated -
X-Zhang/SemTalk-Inference-Data
Updated • 59 -
EchoMask: Speech-Queried Attention-based Mask Modeling for Holistic Co-Speech Motion Generation
Paper • 2504.09209 • Published