yzhao520 / CPP

CVPR 2021 "Camera Pose Matters: Improving Depth Prediction by Mitigating Pose Distribution Bias"

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

CPP module question

Hyeonjeong37 opened this issue · comments

Hello, and thank you for your work!

I'm wondering about the CPP module code.

In line 30, the code below represents equation (5).
p_f = p_f * scale_f.expand_as(p_f) + translation_v

However, z_f is calculated by using k_vec, which is [0,0,1], and the original q with the scale factor.
I can't find where p_f is used in the code, or p_c as well.

If they are not used, why do we obtain them? If I missed something, please let me know.