TheVRMLFaceToolbox
A repo for near mouth blendshape prediction
How does this work?
Its pretty easy, download this repo and run main.py. Currently we are using a test video as input but you can use whatever you want as input!
-
- Add the Vive Model
-
- Decompile SRanipal and figure out how Vive processes the output values from the model. (PROCESS IN REVIEW)
-
- Train a new monocular image model using Vive model. (PROCESS IN REVIEW)
This repo is a work in progress and currently only outputs raw tensor values and not blenshapes