Excited to introduce our new paper, Generative Omnimatte: Learning to Decompose Video into Layers, with the amazing team at Google DeepMind! Our method decomposes a video into complete layers, including objects and their associated effects (e.g., shadows, reflections).

👀👀👀👀👀

Ring deconvolution microscopy is now published at Nature Methods! nature.com/articles/s4159… There are some fun new additions including light-sheet deconvolution 🫡 Stay tuned for the official python package release next week! Any feature suggestions are more than welcome 😃

Yolo just dropped 😂 worth reading till the end theguardian.com/artanddesign/2…

Check out our new preprint! Very fun project co-led with the amazing Hadi Vafaii

PREPRINT THREAD🧵! Ever wanted to know how to ABOLISH THE VALUE FUNCTION, and why IT'S ALL TAXIS NAVIGATION? Maxwell Ramstead and I wrote up some notes for y'all. arxiv.org/abs/2505.17024

Cool useful trick for INRs

Ope looks like Ilya did it too. Add noise to your optimization!

Check out Adobe Project Indigo! Free photography app for iPhone. Cool on-device AI, esp the on-device reflection removal is amazing for taking photos through windows (although it takes a lot of memory, slower for older iPhones) research.adobe.com/articles/indig…

Adobe has quietly rolled out an AI camera app for iOS. Built by legends Marc Levoy (computational photography OG) and Florian Kainz (multi-academy award winner and creator of night sight mode on pixel). Delivers SLR-like image quality with manual controls all on device.

"Perception as Inference" is a century-old idea that has inspired all major theories in neuroscience 🧠 — including: ✅ Sparse Coding ✅ Predictive Coding ✅ Free Energy Principle & more! In my new blog post, I build the intuition behind this idea from ground up 👉[1/6]🧵

Announcing the new "Sensorimotor AI" Journal Club — please share/repost! w/ Kaylene Stocking, Tommaso Salvatori, and Eli Sennesh Details + sign up link in the first reply 👇

Happy to have led local arrangements for #ICCP2025! Massive thanks to our volunteers for making it possible: Dongyu Du, andrew guo, Len Luong, Ali SaraerToosi, Howard Xiao, Andrew Xie, Sophia Yang, and Kelly Zhu! And, Sotiris Nousias and Jazmin Diaz for being the best co-chairs!

Excited to share Flow Matching Policy Gradients: expressive RL policies trained from rewards using flow matching. It’s an easy, drop-in replacement for Gaussian PPO on control tasks.

it always feel funny to ask chatgpt for advice on transformer architectures, like “help me debug your cousin pls”

Excited to share our #CoRL2025 paper: Eye, Robot: Learning to Look to Act 👀🤖 360° teleop data (Insta360) 🎥 → high-speed robotic eyeball 👁️ The robot learns where to look in order to act using a BC-RL perception-action loop. w/ Justin Kerr + team 📎 eyerobot.net

Great talk by Justin Yu presenting the Real2Render2Real paper at CoRL in Seoul! A great collaboration between UC Berkeley and Toyota Research Institute (TRI). The code for this paper is now available, check it out here: github.com/uynitsuj/real2… Toyota Research Institute (TRI) AUTOLab Ken Goldberg

Excited to introduce Dreamer 4, an agent that learns to solve complex control tasks entirely inside of its scalable world model! 🌎🤖 Dreamer 4 pushes the frontier of world model accuracy, speed, and learning complex tasks from offline datasets. co-led with Wilson Yan