search

jixinya / EVP

Code for paper 'Audio-Driven Emotional Video Portraits'.
291 stars 49 forks source link

3DMM model #2

Open prnvjb opened 3 years ago

prnvjb commented 3 years ago

Hi, Awesome work!

I would like know more about 3DMM work. I tried to generate them with a 3dmm model. But didn't get accuracy. Which 3DMM model did you used?

jixinya commented 3 years ago

Hi, I use the BFM09 for geometry and facewarehouse for expression. You can follow the work in face3d.

xiaobairui commented 2 years ago

Awesome

Did you generate it successfully? I got the same problem.

izabalra8 commented 2 years ago

How do you recover the geometry and expression information from 106 facial landmarks? I understand you used the BFM09 and facewarehouse models for that, but I don't find the details of what you mentioned in the paper: "we recover pose, geometry and expression by solving a non-linear optimization problem". Where is that optimization problem stated? Thank you in advance!