[HN Gopher] Sparse Voxels Rasterization: Real-Time High-Fidelity...
       ___________________________________________________________________
        
       Sparse Voxels Rasterization: Real-Time High-Fidelity Radiance Field
       Rendering
        
       Author : jasondavies
       Score  : 111 points
       Date   : 2025-02-21 21:06 UTC (1 days ago)
        
 (HTM) web link (svraster.github.io)
 (TXT) w3m dump (svraster.github.io)
        
       | atilimcetin wrote:
       | I think this paper is as important as original Gaussian Splatting
       | paper.
        
         | mitthrowaway2 wrote:
         | Why do you say so?
        
       | aaroninsf wrote:
       | Can someone ELI5 what the _input_ to these renders is?
       | 
       | 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,
       | 
       | 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!),
       | 
       | but I have no idea what the _input_ is.
       | 
       | I assume if you work in this domain it's understood,
       | 
       | "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...
       | 
       | ...but what I want to know is,
       | 
       | are these video- or still-image input;
       | 
       | and how much/many?
        
         | corysama wrote:
         | > We optimize adaptive sparse voxels radiance field from multi-
         | view images...
         | 
         | Pretty sure the input is the same as for NeRFS, GS and
         | photogrammetry: as many high rez photos from as many angles as
         | you have the patience to collect.
         | 
         | I think the example scenes are from a common collection of
         | photos that are being widely used as a common reference point.
        
         | vessenes wrote:
         | They are photos, in this case from the MIP Nerf 360 dataset. I
         | believe there are on the order of hundreds per scene. They are
         | not videos turned into photos. Some datasets include high grade
         | position and directional information -- I believe this dataset
         | does not, so you need to do some work to orient the rendering
         | training. But, I'm a hobbyist, so all this could be very wrong.
        
       | davikr wrote:
       | What is the usecase for radiance fields?
        
         | corysama wrote:
         | Take a bunch of photos of an object or scene. Fly around the
         | scene inside a computer.
         | 
         | https://news.ycombinator.com/item?id=43120582
         | 
         | Like photogrammetry. But, handles a much wider range of
         | materials.
        
       | loxias wrote:
       | 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.
       | 
       | Love this a whole heck of a lot more than NeRF, or any other "lol
       | lets just throw a huge network at it" approach.
        
         | bondarchuk wrote:
         | > _Love this a whole heck of a lot more than NeRF, or any other
         | "lol lets just throw a huge network at it" approach._
         | 
         | Well yes, but that's what gaussian splatting also was. The
         | question is: are their claims to be so much better than gsplat
         | accurate?
        
           | chpatrick wrote:
           | There's no neural net with gaussian splatting, it's a fancy
           | pointcloud that's optimized with ML techniques.
        
             | bondarchuk wrote:
             | I know, that's the point.
        
               | chpatrick wrote:
               | Mea culpa, I misunderstood.
        
       | bondarchuk wrote:
       | 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.
        
         | ibrarmalik wrote:
         | How is plenoxels a direct predecessor of gaussian splatting?
        
           | yyeboah wrote:
           | They both emerged out of the pursuit of a more efficient
           | solution for addressing the inefficiencies in NeRF, which was
           | mainly due to expensive ray marching and MLP calls. Before
           | the emergence of Gaussian splatting, grids, such as plenoxels
           | were all the rage. Of course, Gaussian splatting here refers
           | to the paper, "3D Gaussian Splatting for Real-Time Radiance
           | Field Rendering"
        
       | HexDecOctBin wrote:
       | 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)?
        
         | yorwba wrote:
         | The reverse-rendering is not real-time, but takes several
         | minutes. Only rendering new viewpoints from the resulting
         | sparse voxel representation runs at high enough framerates.
        
       | markisus wrote:
       | 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.
        
       ___________________________________________________________________
       (page generated 2025-02-22 23:01 UTC)