zhizdev / mvdfusion

[CVPR 2024] MVD-Fusion: Single-view 3D via Depth-consistent Multi-view Generation

Home Page:https://mvd-fusion.github.io/

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

why `n_pts_per_ray=1` in `mvd_train.config`?

thucz opened this issue · comments

In your paper, you said in Sec3.2, "we sample D = 3 depths...", but the actual value in mvd_train.config is 1:
n_pts_per_ray=1.

commented

Indeed, our released model is trained with n_pts_per_ray=3, updating mvd_train.config.

Thanks!