Otto Seiskari

> Another situation where the current approach will definitely barf is when the video and subtitles move at different FPS. This may be easier to solve for in most cases...

Hello! Your findings are consistent with our understanding of the capabilities of the current HybVIO code 1. Sometimes rectification is needed in practice or the left-to-right LK tracking fails (even...

We are not planning to add support for new devices or architectures in this repository. However, if you want to take a shot at fixing the build for the new...

We (Spectacular AI) have also been experimenting with pose optimization with Splatfacto/gsplat and could add the contributions here (I created a new PR https://github.com/nerfstudio-project/nerfstudio/pull/2891 for them since that was technically...

@jb-ye I did not actually check this in detail in https://github.com/nerfstudio-project/nerfstudio/pull/2891/commits/e8e05ff02f1c604124a0373afcca7691f2b88546, and just kept the multiplication order from the [original 3DGS PR](https://github.com/nerfstudio-project/nerfstudio/pull/2521/files#diff-574309e78cc47396f2fd1e783b1695cddbd093a62333ccdfa63bfd659b2a79dfR158) I agree with you that it would be...

> If we want to switch to a single rigid transform, between left- vs right- multiplied my very weak personal bias is actually for right-multiplied, mostly because it seems like...

> I see.. When I activate SO3xR3, I checked the image eval result, the rendered image is little bit shifting due to camera optimization vs GT. I think it was...

Update: * rebased on latest main * fixed linter warnings * exporting optimized poses only for training images To summarize, the following aspects should now be fixed in Splatfacto pose...

> Sorry, Unfortunatelly, it decrease the quality: This is expected. Like I commented here https://github.com/nerfstudio-project/nerfstudio/pull/2885#issuecomment-1936410849, there are two separate cases: 1. The current NeRF/3DGS benchmark datasets which are still images...

> 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...