Martin Ethier

Hello @werner-duvaud, the authors of MuZero have recently published a follow-up paper where they created their own adaptation of MuZero for continuous action spaces (link: https://arxiv.org/abs/2104.06303). Are there plans to...

> Couple things you can try: > > 1. COLMAP the poses to get a sort of upper bound on quality > 2. export the poses from a trained nerfacto...

> Regarding quantitative validation of this PR, Here is my suggestion: > > (1) take a dataset which you think is suitable for pose optimization (2) run a baseline with...

> > I believe depth nerfacto is hooked up to DepthDataset, where if you don't provide any depth data, then ZOE (state of the art monocular depth estimation) depth images...

Same here. It would be better if the authors gave us a link to directly download the dataset (ex: from Google Drive) instead of making us scrape the dataset ourselves.