Invitation to 3d computer vision is highly underrated :)

Well I didn’t speak English very well or basically at all back then, so I wouldn’t know anything about it :p

Ok I might be not up to speed and the pitch itself doesn’t say anything about the game, but what’s going on? I think fondly of both fable and Black and White

These are so gorgeous!

I know that this is not what they're going for, but I find that having their Project Glasswing video open with the CEO of Linux Foundation saying "Most people who use software everyday don't think about bugs" is some next level comedy. I guarantee you Linux users think about bugs. Like all the time

I thought he was taking a break from music? Or was it just touring?

Congrats! Awesome news!

Given how quick the modern phones are, this is not very surprising. I think in general people have not internalized how quick phones are right now - A18 is quite powerful chip, single core perf is matching M1, which people were raving about when it released:
www.cpu-monkey.com/en/compare_c...

Also, thank you, this is awesome!

Wow, big news! Congrats! Slug was very inspiring!

I feel like my very mixed success with LLMs must be a skill issue then.

Any plans for solving the US issue? I guess for original we're SOL, but prints maybe?

I really really hate that the panels on left and right have different look, but it is more of a Liquid glass complain. The tool looks awesome!

Can we please normalize reporting run-time performance in ms not FPS and if you report your time, please also mention the hardware. Pretty pls?

Are you using Claude Code for that, or web interface?

Purely visually reminds me of @keenancrane.bsky.social paper or stripes on surfaces. Not really sure if there is a connection:
www.cs.cmu.edu/~kmcrane/Pro...

glad to hear you're enjoying early parenting

Dear Google Slides, please back off, I like my slide how they are, no need to "beautify" them

Looking at the pdfs in the Github repo, it looks like this framework of Axiomatic Physical Homeostasis can also be applied to solving AGI, which makes me doubly skeptical.

An excerpt from a paper stating that 2011 KinectFusion paper is now an old and traditional method

Wow, thanks for making me feel old, paper

Ok, ok, hear me out.

A PDF reader that doesn't lag.

I know, crazy.

Walk!? In America?? What that activity you speak of?

@parskatt.bsky.social First Dad, now Mum, what are they going to come up with next?

As someone who routinely switches between Linux, MacOS and Windows, the fact that Win11 removed a feature to put taskbar at the top hurts a lot.

Optimized Wireframe Rendering: Part III Detailed technical breakdown of precision pitfalls, clipping logic, and geometry shader implementation for high-quality wireframe rendering.

Finished documenting work on optimized wireframe rendering: precision, clipping & math details for stable, performant edges in OpenGL.

Includes an analysis of NVIDIA’s wireframe paper & fixes its issues.

fauder.github.io/posts/2025/1...

#GraphicsProgramming #Rendering #OpenGL #ShaderDev #GameDev

If we do spatial queries in R^3, should we call the data structure kd-tree, 3d-tree, or just be honest about it and call it sorted pointcloud?

Well, @ericbrachmann.bsky.social did release a hot new paper

Assume that you have a "cone" represented by normal n and cos(θ), where θ is the half angle, how do you compute the intersection of the two without involving acos?
I'm actually mostly interested in the binary classification (interesects / doesn't intersect)

Inside NVIDIA GPUs: Anatomy of high performance matmul kernels - Aleksa Gordić From GPU architecture and PTX/SASS to warp-tiling and deep asynchronous tensor core pipelines.

"Inside NVIDIA GPUs: Anatomy of high performance matmul kernels", includes a great intro to GPU architecture and PTX/SASS: www.aleksagordic.com/blog/matmul

A comparison figure from recent 2025 paper on 3D reconstruction with neural nets. All results are flat shaded pointclouds.

A comparison figure from 2016 paper on 3D reconstruction using RGB-D. This one applies fusion and renders nice meshes

Prove me wrong: For all the 3R papers, there is no reason for not running a fusion over your results. The figures would be way more readable, and misalignment more obvious. The fuzziness of point-clouds hides it.

We have the technology!

2025 vs. 2016