TimoBolkart / voca

This codebase demonstrates how to synthesize realistic 3D character animations given an arbitrary speech signal and a static character mesh.

Home Page:https://voca.is.tue.mpg.de/en

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

Can I train this on custom dataset?

Suvi-dha opened this issue · comments

Hi Timo,

From the paper I inferred that the model can be trained on custom topology based dataset if I could register provided VOCA raw scans on our topology. Am I correct? or the code is strictly dependent on FLAME topology?
If there any other way to train on custom data?