Jake Rice

New at #SIGGRAPH2025: Can we make Perlin Noise stretch along some underlying vector field? Well it turns out it's possible with two simple additions to the original method! No need for advection or convolutions. Find the paper and implementations here: github.com/jakericedesign…

Mesh reconstruction is so well studied, yet new papers still keep coming OffsetAxis shows how, given a UDF, you can convert it to a mesh using medial sphere fitting First, you sample points on an offset surface, creating an oriented point cloud, sort of forming a local shell field Then, you create and optimize medial spheres inside that shell with a variational approach It achieves 2 goals - medial spheres become well distributed along the sampled surface and the duality of the converged state gives you connectivity for free, without an additional triangulation Seems to be better for open surfaces and noisy fields, though it doesn’t provide formal topological guarantees Here is a little doodle of this idea in 2D

MUSK LOSES CLAIMS BARRED BY STATUTE OF LIMITATIONS

@ashlee3dee This is cool!!!!! Have you looked into trivial connections for discrete surfaces? It's a great way to define smooth vector fields on surfaces while prescribing specific singularities :o it's a really neat method imo

learning about harmonic functions for the SplaTool v2 update which will include it as an optional brush stroke direction heuristic

@benclarkson Yeah the fx work is frighteningly solid too. We're definitely getting to the point where it's basically going to come down to storytelling and hoping the audience cares to engage with human made stories and art.

I think the AI on this looks, frankly, pretty good. I am not prejudiced since many accuse my best work of being AI. However the storytelling is absolute crap. Visually there's no clear story. It's a "trailer" - "I'll tell the story later" - sorry bruh, you had two minutes.

📢📢📢introducing 𝐏𝐨𝐰𝐞𝐫 𝐅𝐨𝐚𝐦 A 3D representation that can be ray traced or rasterized in real time, with NO COMPROMISE in quality. - Project: powerfoam.github.io - arXiv: arxiv.org/abs/2604.24994 Rasterized at 3DGS-class FPS Ray traced at Radiant Foam speeds

This January, I decided to give it a shot and wrote my first paper Today, I am happy to share that it was accepted by #SIGGRAPH2026 SAD is a differentiable image representation with soft, anisotropic partitioning, with up to 20x faster encoding time🧵 luckyiyi.github.io/SAD/index.html

I made an experimental "GPU mode" in my toy shader debugger (pema.dev/hlsl/). Runs the shader on the GPU with WebGPU by transpiling to WGSL with the Slang compiler. You can click on a pixel to drop into the interpreter and start debugging. Also added image previews.

Got accepted to SIGGRAPH 2026! My talk is titled: Reducing graphics API complexity: A clean slate API design for modern hardware I am an ARM Ambassador nowadays. I am going to be talking in their track: developer.arm.com/community/arm-…

3D position-based blur in COP, no UV seams. 🎨 - 2 OpenCL nodes. - 1st one stores positions in a quadtree. - 2nd one blurs the image using it. #Houdini

x.com/i/status/20077… Extending Gabor kernels to neural implicits makes a lot of sense. Super cool results! I did a similar test but with gabor kernels and soft voronoi!!

Make no mistake: destroying world-class universities, like the US just did with Sharif University of Technology in Tehran, isn't just an attack on Iran but it's literally an attack on all of us, on all of humanity. It's not Iran that "won" when Maryam Mirzakhani made her discoveries that won her a Fields Medal: it's all of mathematics, and everything mathematics is used for. Human progress won, technology won, we all won. It's the same type of stuff the Mongols did during the sack of Baghdad and their destruction of the House of Wisdom: we ALL lost something irreplaceable back then, entire fields of human knowledge set back. That's what bombing a university does. It doesn't just destroy buildings. It destroys us, all of us.

@yiningkarlli I think being slightly more explicit with it would probably help tie it together. I mean I think the true problem of this theory is how would someone know they have enough domain knowledge to use AI? The chat bots love to make you feel like you're already at that level.

@TearsOfJake I assumed that Alice in the end does as well but is now able to use it better than Bob can.

I find the resolution to the Bob and Alice narrative at odds with a lot of the messaging of the essay. The author espouses that using AI is effectively fine once you've paid your dues to build intuition. Yet in the fictional world, Alice never does? But otherwise I enjoyed this!

先週ラボを博士卒業した@tojo_kenji さんの最新作(CVPR2026)です。圧倒的に少ない三角形で高品質な再構築。既存手法にありがちな深度ソートが不要で、シンプルなシェーダで超高速に描画可。鍵となるのは離散的なアルファ値を微分可能にする新たな確率的学習法です　 youtube.com/watch?v=xoogpi…

@MattiaMerenda2 The new OTIS solver in Houdini uses VBD I believe. So it's already happening :)

@TearsOfJake I wonder how fast we will see wide use of a successor to PBD, something like Augmented Vertex Block Descent showed by Utah Uni seems extremely promising in usually high iteration scenarios, still in awe at your greeble setup btw

The perils of discretization

Our #CVPR2026 work Differentiable Straightest Geodesics [circle-group.github.io/research/DSG/] provides exponential maps for 3D meshes allowing to trace geodesics or parallel-transport vectors. With it, we release DiGeo, a PyTorch differential geometry package: digeo.readthedocs.io. 👇

VoroNERF is coming along now that i've fixed the problems with my top-k selection of voronoi sites! Still not perfect, but by far my best attempt yet. This is using 60K voronoi sites, and a top-k of 16 voronoi sites at any given sampling position along a ray.

@z8editing So like, I don't think it's important that people know that I spent months of my life, simulating hedges on Zootopia 2. What I do think is important is that audience feels like the result was intentional. That there was effort put in to give them something unique.

@z8editing Does the average person know what goes into a good piece of photo-bashed concept art? I used the word effort, but what im trying to convey is the perception of effort and by extension, intention. Belief that the artist cared enough to be intentional with the final product.

Anime is actual a great place to base this discussion. I did a quick search and found thousands of posts talking about CGI in anime being lazy. Clearly people care about perceived effort. To anime viewers CGI is considered low effort, and it hurts the experience.