Zirui Wu
> js里的授权域名失效了，换一下就好。 可以说的详细一些吗，谢谢！
It seems(?) NeRFStudio does not support using multiple viewer methods (e.g using tensorboard and viewer at the same time) currently. Will it be a future feature?
Met the similar problem, seems like during training the dataloader forces images to be in landscape (https://github.com/facebookresearch/vggt/blob/training/training/data/datasets/co3d.py). Is this true for all datasets?
> Hi, do you mean the arrow key controls? > > > > If so the velocities are unfortunately hardcoded, you could adjust them here: > > https://github.com/nerfstudio-project/viser/blob/b947d2480710e8d3bfaadceea8ea921e28011675/src/viser/client/src/CameraControls.tsx#L133-L192 Thanks for...
Yes, it could also be a proper solution, I totally agree.
> Hi @maturk! Thanks for the reply, but if this is the case, the grad value for those gaussians would be zero instead of nans? For my data, the training...
> hmm yes that is strange, what kind of dataset are you trying to train? I am wondering if there are any problems with projection. @wuzirui, you can check the...
> If the scales is super small, it could be numerically instable. Maybe try adding a small eps to the scales? > > ``` > scales = torch.exp(scales_crop) + 1e-4...
> Hi @wuzirui, did you figure out your problem? not yet, maybe later I need to look into the kernel
Hey guys, I've solved the issue by simply scaling up the scene by a constant number: all the location-related values (i.e. means, translation part of the camera pose...) ---- the...