James Perlman

Happy 2025! Earlier last year I stumbled upon NVIDIA Warp, a differentiable, kernel-based framework that lets you write CUDA-like code in python. I've been making a new NeRF app with it and have found it to be extremely performant and fun to use! Enjoy this short training video. This is Warp + PyTorch + fVDB + tcnn. Enjoy!

@taiyasaki This is so scary and my heart goes out to you & your family & your baby. Get well soon ❤️‍🩹

Afraid I am cancelling my #CVPR2025 trip… Newborn has an infection and admitted to the paediatrics ward. Enjoy Nashville

There is value in your uniquenesss

@RVg80361301 It’s for doing differentiable raymarching and being able to control how raymarching is done with python

@jperldev Hi Sir, what was the purpose of Warp for this application? :)

Happy 2025! Earlier last year I stumbled upon NVIDIA Warp, a differentiable, kernel-based framework that lets you write CUDA-like code in python. I've been making a new NeRF app with it and have found it to be extremely performant and fun to use! Enjoy this short training video. This is Warp + PyTorch + fVDB + tcnn. Enjoy!

@RadianceFields Damn right it is!

RadFoam is out with the same license as Nerfstudio!

If you’re into raytracing, this is the most exciting radiance field representation since hash grids. I feel loved. Happy Valentine’s Day! 💞

@BenR82225 Splatting gaussians offers better rendering framerates for the most common hardware currently, that is the benefit. Besides that, NeRFs tend to take up less memory and can have higher realism

@jperldev Oh why that? I thouth they where more advanced then some nerf systems

This is a Foliage-emulating Radiance Network, or FeRN for short. Ok just kidding it's just a NeRF of a fern I spotted on a hike but it's my post for the day. Enjoy. #WarpNeRF

@natanielruizg I had this issue too and after manually using the "I don't want to see this" option on a bunch of posts I'm back to academic research feed!

if you want to maximize unregretted minutes spent on this app why am i getting viral tiktok videos instead of AI research posts in my feed

@leondenise This is the hardest part.

I should probably share the addon already before I step in the endless vortex of features creeping and dive in the never ending tunnel of "it's not ready before it's perfect"

@leondenise THIS IS SICK

Because I don't have enough problems in my life, I decided to make a custom render engine in #Blender so I can code GLSL shaders and render them nicely

@BenR82225 Gaussians are boring

@jperldev Haven't heard anything about nerfs for a while since gausian spalts are a thing

The motto on this arcade game machine perfectly encapsulates my feelings on radiance fields.

@3pointedit This is so crazy they’ve just built an office under the water line. The forces holding back instant destruction must be astronomical.

Hard to believe this isn't the backlot for Dune. 😲

Siren vibes. NeRF quality improvements are on hold for now while I work on the utterly ridiculous idea of implementing a remote websocket renderer for use in a custom Blender addon. Will it work? Probably! Will it be fast? Hell no! #WarpNeRF

@RadianceFields This looks very radiant to me!

It really looks like there’ve been two radiance fields in Jensen’s keynote so far

@mazy1998 Yes! I haven’t announced an official release yet, but the code is visible at github.com/JamesPerlman/W… Be forewarned - it is entirely undocumented.

@jperldev Will you release the code for this? I’m interested in warp

here you go, good ol' slerpin quats.

Balloon Dog by Jeff Koons at the Broad Museum, Los Angeles. Continuing with static images until I can get camera movements working in the Blender addon/renderer.

NeRF for now, because the raymarching algorithm has a ton of untapped artistic potential to create some really unique stuff. I’ve worked with gaussians and they have a lot of benefits over NeRF, also I’d like to experiment with raytracing them, but splatting them isn’t too useful to my goals.

@jperldev Have you been doing much with GS yet or primarily NeRF?

#WarpNeRF - Posting a NeRF a day until I get a job😄 10k steps usually trains in under 60s @ 4096 rays/sec (largely thanks to tiny-cuda-nn) and this can be improved much further. Currently working on a Blender addon to control the camera, then will fix the image quality :]

Portrait of two fashionable desert nomads, trained in WarpNeRF. Without a contraction function, background collapse causes blurriness in scenes with a large amount of depth.