TimoBolkart / voca

This codebase demonstrates how to synthesize realistic 3D character animations given an arbitrary speech signal and a static character mesh.
https://voca.is.tue.mpg.de/en
1.14k stars 271 forks source link

Assign random shape parameters #93

Closed yangdaowu closed 2 years ago

yangdaowu commented 2 years ago

Hello, I want to consult the betas [: 100] is how to set up.

for i in range(num_samples):

Assign random shape parameters.

# Beware changing expression parameters (i.e. betas[300:400]) or pose will result in meshes that cannot be used
# as template for VOCA as VOCA requires "zero pose" templates in neutral expression
flame_model.betas[:100] = np.random.randn(100) * 1.0

# Output mesh in "zero pose"
Mesh(flame_model.r, flame_model.f).write_ply(out_fname(i+1))
TimoBolkart commented 2 years ago

What do you mean with you want to consult the betas? What do you want to do?

TimoBolkart commented 2 years ago

The edit_sequences script demonstrates how to alter face shape of a sequence, if this is what you intend to do.