You couldn’t ask to work for a nicer guy or more talented team!

Interesting connection between ML attention (which is basically a data-dependent weighted average) and the bilateral filter from image processing…

I'm very happy to announce that our paper "Don't Splat your Gaussians: Volumetric Primitives for Rendering Scattering and Emissive Media" (tinyurl.com/GaussVol) was finally accepted to ACM Transactions on Graphics last month! We will present it at SIGGRAPH 2025 🧵🧵🧵 (1/11)

I like the new linear swept spheres, looking forward to somebody using that as a primitive for scene reconstruction!

Worth watching! It’s impressive how well Charlie Brooker extrapolates current tech. Walking into “3d” versions of old photos isn’t *that* far from being possible with generative AI, though maybe not with brain interfaces…

[SIGGRAPH '25] Monocular Online Reconstruction with Enhanced Detail Preservation Abstract (excerpt): Our approach addresses two key challenges in monocular online reconstruction: 1. Distributing Gaussians without relying on depth maps. 2. Ensuring both local and global