zhengyuf / IMavatar

Official repository for CVPR 2022 paper: I M Avatar: Implicit Morphable Head Avatars from Videos
MIT License
633 stars 61 forks source link

The size of the output mesh is too small #18

Closed 1208869644 closed 2 years ago

1208869644 commented 2 years ago

thank you for your great work,by using the unsupervised pretrained model, the output mesh is only 2.4M. Is there any way to improve the output mesh?

zhengyuf commented 2 years ago

You can change the res_init and res_up in the .conf file to eliminate the marching cube artifacts, but then it will be slower.

Zvyozdo4ka commented 8 months ago

@1208869644 how did you get mesh?

@zhengyuf is it possible to get point cloud?