I look forward to reading this in closer detail, but it looks like they solve an inverse problem to recover a ground truth set of voxels (from a large set of 2d images with known camera parameters), which is underconstrained. Neat to me that it works w/o using dense optical flow to recover the structure -- I wouldn't have thought that would converge.<p>Love this a whole heck of a lot more than NeRF, or any other "lol lets just throw a huge network at it" approach.
This is basically Gaussian splat using cubes instead of Gaussians. The cube centers and sizes choices are discrete and non overlapping, hence the name “sparse voxel”. The qualitative results and rendering speeds are similar to Gaussian splat, and it’s sometimes better or worse depending on the scene.
Why is this called rendering, when it would be more accurate to call it reverse-rendering (unless "rendering" means any kind of transformation of visual-adjacent data)?
Funny, it almost sounds like a straight efficiency improvement of Plenoxels (the direct predecessor of gaussian splatting), which would mean gaussian splatting was something of a a red herring/sidetrack. Though I'm not sure atm where the great performance gain is. Definitely interesting.
Can someone ELI5 what the <i>input</i> to these renders is?<p>I'm familiar with the premise of NeRF "grab a bunch of relatively low resolution images by walking in a circle around a subject/moving through a space", and then rendering novel view points,<p>but on the landing page here the videos are very impressive (though the volumetric fog in the classical building is entertaining as a corner case!),<p>but I have no idea what the <i>input</i> is.<p>I assume if you work in this domain it's understood,<p>"oh these are all standard comparitive output, source from <thing>, which if you must know are a series of N still images taken... " or "...excerpted image from consumer camera video while moving through the space" and N is understood to be 1, or more likely, 10, or 100...<p>...but what I want to know is,<p>are these video- or still-image input;<p>and how much/many?
Reminded me of the Radian Foam article posted here[1] not long ago, though the focus there was on being differentiable.<p>[1]: <a href="https://news.ycombinator.com/item?id=42931109">https://news.ycombinator.com/item?id=42931109</a>