41–50 of 90 posts
I'm still highly skeptical of gaussian splatting as anything more than a demo. The files are too large. The steak sandwich is 12meg (as just one example) There was a guassain splat based Matterport port clone at least year's siggraph. To view a 2 bedroom apartment required streaming 1.5gig Cool demo
I'm still highly skeptical of gaussian splatting as anything more than a demo. The files are too large. The steak sandwich is 12meg (as just one example) There was a guassain splat based Matterport port clone at least year's siggraph. To view a 2 bedroom apartment required streaming 1.5gig Cool demo
Thanks! Notice 12MB steak sandwich is the biggest of them all. Rest are Fancier compression methods are coming (e.g SOGS). This is 30MB! https://vincentwoo.com/3d/sutro_tower/
I'm still highly skeptical of gaussian splatting as anything more than a demo. The files are too large. The steak sandwich is 12meg (as just one example) There was a guassain splat based Matterport port clone at least year's siggraph. To view a 2 bedroom apartment required streaming 1.5gig Cool demo
Thanks! Notice 12MB steak sandwich is the biggest of them all. Rest are Fancier compression methods are coming (e.g SOGS). This is 30MB! https://vincentwoo.com/3d/sutro_tower/
Would the format work better if you made that cut-off at something like 1 sigma instead? Then instead of these blurry blobs you'd effectively be rendering ovals with hard edges. I speculate out loud that maybe you could get a better render with fewer hard-edged ovals than tons of blurry blobs.
We seem to have to poles: extreme realism, and extremely minimalistic pixel art. I prefer the second camp. But your project looks really important in the first camp.
Do you have any insights into the current performance bottlenecks? Especially around dynamic scenes. That particle simulation one seems to struggle but then improves dramatically when the camera is rotated, implying the static background is much heavier than it appears.
And as a counterpoint to the bottlenecks, that Sierpinski pyramid, procedurally, is brilliant.
any chance this will support the video style 3d gaussian splats?
Very very cool. Do you have any insights into the current performance bottlenecks? Especially around dynamic scenes. That particle simulation one seems to struggle but then improves dramatically when the camera is rotated, implying the static background is much heavier than it appears. And as a counterpoint to the bottlenecks, that Sierpinski pyramid, procedurally, is brilliant.