foivospar / ned Goto Github PK

Thank you very much for doing such an excellent job. I can ask where the video materials for training in the paper can be obtained, and I cannot open the link for contact information here.

Can the method be tested directly on images (not videos)? For example, two images are used to test expression transfer.

How long did it take you to train a new actor and what GPU specs

Hello, I used a new actor video and I trained a new neural face renderer, but the mouth of the generated mp4 video is blurred and cannot be overlapped with the original. Is this related to the video I used of Asians or something wrong with my testing or models? In addition, could you give me your email? I can't open the email link in github and I want to contact you to apply the training/test videos used in the experiments, thanks a lot.

Hey @foivospar hats off to the great work. I wanted to ask is there any tutorial and easy steps by using which we can run it on windows. Any help in this regard will be very much helpful.

训练时， 运行python preprocessing/segment_face.py时报错，
from postprocessing.image_blending.image_blender import SoftErosion，但是看文件中并没有image_blending这个文件夹

Possible to edit the expresion of the 3D model in flame or blender or something?, I need to keep the mouth closed.