Slides - Video
You can find the code of our approach published in Computer Vision and Image Understanding (CVIU) at:
https://github.com/clferrari/deep-3dmm-refinement