This repository contains the code for the CVPR 2020 paper "Differentiable Volumetric Rendering: Learning Implicit 3D Representations without 3D Supervision"
Hi, thanks for this great work and your other inspiring work in reconstruction!
I want to add more local information to condition code c. Thus the projection from 3D point in world space to 2D location in input image is needed.
I found the function transform_to_camera_space could transfer the p_world to p_cam by providing three camera mat. I want to confirm that if the p_cam[:, :, :2] is the projected 2D location(in input image) of p_world given input_camera_mat.
Hi, thanks for this great work and your other inspiring work in reconstruction!
I want to add more local information to condition code c. Thus the projection from 3D point in world space to 2D location in input image is needed.
I found the function
transform_to_camera_space
could transfer the p_world to p_cam by providing three camera mat. I want to confirm that if the p_cam[:, :, :2] is the projected 2D location(in input image) of p_world given input_camera_mat.Thanks!