search

theEricMa / DiffSpeaker

This is the official repository for DiffSpeaker: Speech-Driven 3D Facial Animation with Diffusion Transformer
137 stars 18 forks source link

Blendshape generation #14

Closed mahekchandak closed 3 months ago

mahekchandak commented 3 months ago

How do I get the ARKit 51 blendshapes value from the video?

theEricMa commented 3 months ago

Hi, this work is focused on generating speech-driven 3D facial coefficients. I believe there are open-source repositories or apps that allow for ARKit capturing.