1 paper
Hao Li, Ju Dai, Xin Zhao +3
In 3D speech-driven facial animation generation, existing methods commonly employ pre-trained self-supervised audio models as encoders. However, due to the prevalence of phonetical…