WildGaussians: 3D Gaussian Splatting in the Wild

Not only is there code but they&#x27;ve actually pinned version numbers in their requirements.txt!<p>Truly an outlier ;)

Corridor Digital created a short film and accompanying making-of video showing the use of gaussian splats vs. photogrammetry to reconstruct an environment: <a href="https:&#x2F;&#x2F;youtu.be&#x2F;GaGcLhhhbDs?si=rJVxF8yNwwbfLBw9&amp;t=306" rel="nofollow">https:&#x2F;&#x2F;youtu.be&#x2F;GaGcLhhhbDs?si=rJVxF8yNwwbfLBw9&amp;t=306</a>

The ability to re-light scenes is a step in the right direction, but their examples use the classic easy scenario of an overcast sky with no strong lighting directionality, which can be faked fairly convincingly using ambient occlusion rather than fully simulating (or inferring) the light paths.

video here: <a href="https:&#x2F;&#x2F;wild-gaussians.github.io&#x2F;" rel="nofollow">https:&#x2F;&#x2F;wild-gaussians.github.io&#x2F;</a>

I really want them to create a mascot image for the wildgaussian project that is a T-Rex holding two NErF blasters.

Any insights on what it does better than the concurrent works (Gaussian in the Wild, SWAG, Wild-GS)?

I get my NERF&#x2F;Guassian etc updates from <a href="https:&#x2F;&#x2F;x.com&#x2F;janusch_patas" rel="nofollow">https:&#x2F;&#x2F;x.com&#x2F;janusch_patas</a> This guys posts researches only about these things. I was also going to submit wild guassian.

Can someone eli5 splatting? How does it differ to normal photogrammetry techniques