search

yfeng95 / PRNet

Joint 3D Face Reconstruction and Dense Alignment with Position Map Regression Network (ECCV 2018)
http://openaccess.thecvf.com/content_ECCV_2018/papers/Yao_Feng_Joint_3D_Face_ECCV_2018_paper.pdf
MIT License
4.96k stars 944 forks source link

Texture Problem #3

Closed BouOus closed 6 years ago

BouOus commented 6 years ago

Why don't I find good results in terms of texture?

saved

I changed the code to set opencv to True (for texture extraction) but nothing changed

Is this due to meshlab? i have joined an output example

0.zip

Thank you

yfeng95 commented 6 years ago

Yes, meshlab renders the mesh by default. You can close the light by Render->Lighting->Light off (see the screenshot) to view the original color. meshlab

BouOus commented 6 years ago

Thank you very much, i have solved the problem ;)

Another question please, Is there a way to retrieve the code for training ?

yfeng95 commented 6 years ago

If you mean the code for training the regression network, you can offer your email address then I can send you the code. It's very easy since I provide the network code for testing. However, most of my work are preparing the training data, if you also need the code for generating the UV position map, there are some problems. One problem is that my code for generating training pairs is chaotic now, and some are written in C++ and compiled to python lib, I am not sure whether it can work at other machines. The other is that I have modified many other 3D data(BFM, expression, UV coordinates,etc) into my own version for better usage and I am afraid I have no permission to make it public. So sorry for that.

BouOus commented 6 years ago

Thank you very much, I sent you an email