A question about the 3.2Section of the paper

autonomousvision / differentiable_volumetric_rendering

This repository contains the code for the CVPR 2020 paper "Differentiable Volumetric Rendering: Learning Implicit 3D Representations without 3D Supervision"

http://www.cvlibs.net/publications/Niemeyer2020CVPR.pdf

MIT License

794 stars 90 forks source link

A question about the 3.2Section of the paper #71

Open sleep2hours opened 2 years ago

sleep2hours commented 2 years ago

Hello! I wonderded this equation when I read this. According to the multivariate chain rule, should it be this: XL$EWNKE2`{F6JUD`Z3BG~D And if so, how do you derive this section: W3F7X }FEO$G$4CE{1Q0SC

sangminkim-99 commented 1 year ago

The Eq 5. is changed as below.

Since $\hat{p}$ also depends on $\theta$, you need to use total derivative of $\hat{I}_u$. Reference: https://en.wikipedia.org/wiki/Total_derivative

Thank you.