search

barbararoessle / dense_depth_priors_nerf

Dense Depth Priors for Neural Radiance Fields from Sparse Input Views
MIT License
389 stars 49 forks source link

On the standard deviation of depth map #20

Open joeyw1030 opened 2 years ago

joeyw1030 commented 2 years ago

std = (((z_vals - depth.unsqueeze(-1)).pow(2) * weights).sum(-1)).sqrt() I don't quite understand the concept of standard deviation on your side. Can you expand it?

barbararoessle commented 1 year ago

This implements a weighted sample standard deviation, i.e., the square root of the weighted sample variance.

z_vals are the depths of samples along the ray depth is the weighted mean of z_vals weights are the normalized weights of the samples