jiaxiangshang / MGCNet

Self-Supervised Monocular 3D Face Reconstruction by Occlusion-Aware Multi-view Geometry Consistency[ECCV 2020]

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

Questions about the pose and projection matrix in the code

deepmo24 opened this issue · comments

Hi, sorry to disturb you. I have several questions about the pose and projection matrix code.

1. why we need to add defined_pose_main to pred_pose_render?
image

2. what do the variables "list_gpmm_mv" and "list_gpmm_eye" do ? I found they are calculated from rotation matrix and translation vector, and used in mesh_renderer_camera_light(), but I don't know the meaning of them.
image

3. in project3d_batch(), why the projected landmarks need to be divided by 'pt_batch_homo_2d_w'
image

The basic questions, I suggest you ask your advisior.

Yes, these are actually basic questions hhhhh... I will figure them out. Thanks.

For the first question, this has an insight that gives a main_pose can help the network convergence fast, it just like give a good init value.

thank you so much.