A 3D facial animation framework that disentangles content and emotion and predicts frame-wise emotion intensity from audio plus text for dynamic expressions.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CV 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
MEDTalk: Multimodal Controlled 3D Facial Animation with Dynamic Emotions by Disentangled Embedding
A 3D facial animation framework that disentangles content and emotion and predicts frame-wise emotion intensity from audio plus text for dynamic expressions.