samples in inverse depth or Euclidean space?

[cv-lab-x]

hi, thanks for your great work. I have some questions about samples in nerfacc, in mipnerf360, they do uniform samples in s-space(inverse depth), and transform sample pts from s-space to t-space(Euclidean space). sdist = helper.sample_intervals( randomized, sdist, logits_resample, num_samples, single_jitter=self.single_jitter, domain=(init_s_near, init_s_far), ) tdist = s_to_t(sdist)

while in nerfacc ray_marching.cu, did you do samples in s-space or directly uniform samples in t-space(Euclidean space)?

looking forward to your reply @liruilong940607

[liruilong940607]

It is directly sampled from t-space. We followed the way of sampling from NGP.