A causal autoregressive model for personalized audio-driven facial animation uses a temporal hierarchical motion representation and a multi-modal style retriever to achieve low-latency adaptation without pre-encoded templates or audio lookahead.
InProceedings of the IEEE/CVF conference on computer vision and pattern recognition
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.GR 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Personalizing Causal Audio-Driven Facial Motion via Dynamic Multi-modal Retrieval
A causal autoregressive model for personalized audio-driven facial animation uses a temporal hierarchical motion representation and a multi-modal style retriever to achieve low-latency adaptation without pre-encoded templates or audio lookahead.