Question about SparseNeuS metrics

[JamChaos]

Hi, nice work！
Have you trained SparseNeuS on the Re10K/scannet dataset? I noticed that the reported metrics in the paper are quite low compared to other methods.
Thanks a lot for your time!

[HanzhiChang]

Hi! Thanks for your attention and sorry for my late response.
We retrained SparseNeuS on Re10K & Scannet datasets for 150k (stage 0) + 150k (stage 1) steps according to its original settings. Because the original code does not support for these datasets, we reimplemented it on our own. We have double-checked our code and retrained for another time recently. The metrics are still much lower than other methods.
Thanks a lot!