Hoang-Son Vo
Browse
A two-stage Mamba model that predicts fMRI brain activity from images by going coarse-to-fine, and along the way finds causal evidence that it splits into an early-visual-cortex pathway and a semantic pathway.
Audio-driven talking head generation presents significant challenges in creating realistic facial animations that accurately synchronize with audio signals. This paper introduces ATL-Diff, a novel approach that addresses key limitations in existing methods through an innovative three-component framework.
© Hoang-Son Vo 2022 - 2026