Can use just 10 images to teach a model what a sea anemone looks like?
In our #neurips2026 paper, "Dataset Distillation via Drifting," we distill ImageNet's 1,300 images per class down to just 10, optimized from noise.
https://t.co/NBnFsVpkJ8
More info in thread! ⬇️
1/N
Glad to share that World from Motion has been accepted to #NeurIPS2026! We’re releasing the code, model weights and DyCheck reproduction scripts:
Code: https://t.co/q3cpUp3YCz
Model weights: https://t.co/MuYVlz2GDP
See you in Sydney 🌏🇦🇺🐨!
#neurips2026#NVIDIA@alextrevith
🚀 Introducing ATAS: Any Time, Any Space.
Every shot has a history.
Video agents share a persistent world state to revisit places and replay events from new viewpoints.
Explore the interactive demo: https://t.co/SviUw7BRIM
I am offering a new course on Principles of Deep Representation Learning at HKU based on our new open-sourced book: https://t.co/leZlkURb7j The first six lectures (in English) have been recorded and made available on Bilibili: https://t.co/va95itVovv New lectures will be released there too.
Was meant to be a Valentine's gift, but I'm releasing it early ❤️
Wrote a long piece on KV caching in flow models for image generation.
Includes intuition, commentary, implementation details, benchmarks, & some quirky experiments.
https://t.co/Ufrp5jZIf1
Added arrows for the 3 easiest passing lanes, each with a completion % estimate, on real @MLS game footage in the 3D @Apple Vision Pro replay ⚽️
Used Claude Opus 5.5 to:
• track every player + the ball (RF-DETR, BoT-SORT, DINOv2)
• train two gradient-boosted pass models (un-xPass method) on thousands of real Bundesliga passes
#soccer #futbol #visionos #vr #ar
Let's remove VAEs and ViTs from video models!
🚀 Introducing 𝗣𝗶𝘅𝗲𝗹𝗨𝗠𝗠: an encoder-free unified multimodal model for image and video understanding and generation, directly in pixel space.
Code and model available today!
🌐 https://t.co/RXUObhsSIZ
📄 https://t.co/lA7VaJoVws
LLMs can already build a Blender scene from a room video. So is 3D vision doomed?
We tried the opposite: hand the vision models to GPT-6 as tools (Pi3X for geometry, SAM3, GVHMR/Kimodo for people). It turns a room video into a better interactive sim.
👇
https://t.co/UhYUyNaNAS
3D/4D reconstruction is one of the great achievements of computer vision. With roots in photogrammetry, multi-view geometry was much studied (e.g. Hartley & Zisserman). In the last decade, it was successfully combined with deep learning. Shape priors which come from recognition got incorporated (e.g. for humans, in HMR) and recent systems like VGGT and SAM 3D can do amazing stuff. This progress is directly relevant for robotics, which (duh) operates in the 3D world. Many robotics papers in the learning era weren't exploiting 3D structure, which IMHO is just wasting valuable signal. But there has been a strong real2sim2real tradition for years e.g. here are two projects from my group (and there are many others, I am not claiming exclusivity). I am glad that people have rediscovered this capability with Astra. But please do give credit to human researchers, not just AI models. https://t.co/OeBSOfO6Sg ,
https://t.co/7ZIrP8SPU1
Jev is one of the things that, if the concept proves its worth, I believe would bring a revolutionary shift in AI and Gen AI applications in terms of speed and huge cost reduction.
This is a video that explains Jev in a simpler way.
Introducing AirVis Spatial:
Bring AI to the physical world.
Let AI explore your 3D scans, move through a scene, and look around from different perspectives. Understand what’s there, where things are, and how they relate to one another.
Available now for free with AirVis Studio (https://t.co/LKusftP9tH)
Three simple steps with AirVis Studio:
Record a video → Create a 3D space → Connect to Codex.
Credits:
Lublin scan by Andrii. Film clip from Mercy.
"An Introduction to Flow Matching and Diffusion Models" is a set of MIT lecture notes from the 2026 course "Generative AI" that I found useful for understanding the mathematics behind modern generative AI.
The notes build the subject gradually, starting from probability distributions, ordinary and stochastic differential equations, and Brownian motion, before moving to diffusion processes, flow matching, score matching, and classifier-free guidance. They also explore architectures for image and video generation, latent spaces, autoencoders, and discrete diffusion models for language generation.
It's a really interesting resource.
https://t.co/J96rHCBPrb
World Labs co-founders Justin Johnson and Dr. Fei-Fei Li say Atlas unifies the two things computer vision has always kept apart:
Justin: "Historically, reconstruction has been its own subfield in computer vision with its own specialized tasks and models. Generation is what all the text-to-video models are really good at."
"Those are great for creative applications if I want to imagine something that's never been there before."
"But now with Atlas, for the first time, we're putting these two different parts of visual intelligence together in one model. It can do both 3D reconstruction and generation together in one architecture."
"We had to make it multimodal from the start. This thing natively works on text, images, videos, and camera poses as a native input to the model... It uses 3D as a native modality that it works on."
Fei-Fei: "Computer vision has been around for more than half a century... Our field traditionally has multiple tracks. You go to a computer vision conference: you have the pixel generation track, you have some recognition track, and you have a 3D reconstruction track."
"This is an elegant model that unifies the problem of reconstruction and generation by anchoring on viewpoints and viewpoint estimation. That's just incredibly powerful."
@jcjohnss@drfeifei
A compressed ViT shrank from 327.42MB to 6.01MB, a 54.5x cut, while matching 95.13% FP32 on a cross-village chilli test. A same-size INT8 student trained directly reached 94.87%. The deployment win is clear; the distillation gain remains unshown.
https://t.co/mtymmwC6Kl
48 consistent camera feeds of me all from one casual video capture. This is 4DAnyone. I am testing it out over the holiday week, the goal is 4D Gaussian Splatting from a single camera.
Should out to @pablovelagomez1' for inspiring me to use Rerun to visualize my progress. Right now you see 6 cameras, but I can cycle through all 48.
I hacked together this visualization using ChatGPT's new Astra model in VS Code. Literally took less than 10 minutes.
More coming soon.
Learn more about 4DAnyone Here: https://t.co/bDx9o5brg8
You can also try Pablo's 6 view 4DAnyone generator on Hugging Face here: https://t.co/8jQ1F4oFht
@jeremyparkphd@Scobleizer
#Computervision #3D #4DGS