Alex Bewley

@moo_jin_kim Interesting. How does it compare to [2512.15692] mimic-video: Video-Action Models for Generalizable Robot Control Beyond VLAs share.google/UW7DOfBCVbgVkw… ?

We release Cosmos Policy 💫: a state-of-the-art robot policy built on a video diffusion model backbone. - policy + world model + value function — in 1 model - no architectural changes to the base video model - SOTA in LIBERO (98.5%), RoboCasa (67.1%), & ALOHA tasks (93.6%) 🧵👇

@ducha_aiki Are the LLM sounding reviews generally more positive or negative in their overall ratings?

I am reading now the reviews by fellow reviewers (I didn't submit for CVPR) in my batch, and, OMG, I see so many ChatGPT reviews.

We’re bringing powerful AI directly onto robots with Gemini Robotics On-Device. 🤖 It’s our first vision-language-action model to help make robots faster, highly efficient, and adaptable to new tasks and environments - without needing a constant internet connection. 🧵

Last week, we joined as co-sponsor of the 7th Robot Learning Workshop at #ICLR2025 in Singapore - themed "Towards Robots with Human-Level Abilities". Here's a recap of what went down 👇🧵 1/10

Having organized the robot learning workshop for a number of years now. I'm ecstatic at how engaged the ICLR community has been for its inaugural appearance at #ICLR

@SepehrSamavi @IeeeTro @florian_shkurti @angelaschoellig Nice work in using trajectory forecasting for safe navigation around people. Have you evaluated your predictions on the JRDB benchmark? I didn't see it in your paper. There you may also find some open source trajectory prediction methods to try with your MPC planner.

Our latest work Safe and Interactive Crowd Navigation (SICNav) has been accepted to @ieeeTRO! SICNav jointly solves planning and prediction while satisfying safety consts! Thanks @florian_shkurti @angelaschoellig preprint & code: sepehr.fyi/projects/sicnav youtu.be/BY23UXwM1sM

🚨 Alex Bewley presenting now “Robots and Understanding the World around Them”! #ECCV2024 #OpenSUN3D @eccvconf 📍 Room: Amber 4 🌍 Schedule: opensun3d.github.io

Introducing our Keynote Speakers at this edition of the OpenSUN3D workshop #ECCV2024 (Sept 29, Sunday 14:00-15:30, Room: Amber 4) in Milano🇮🇹 Full schedule: opensun3d.github.io 🚀 @eccvconf @ETH_en @ETH_AI_Center @Stanford

Meet our AI-powered robot that’s ready to play table tennis. 🤖🏓 It’s the first agent to achieve amateur human level performance in this sport. Here’s how it works. 🧵

Open X-Embodiment wins the Best Paper Award at #ICRA2024 🎉🤖! An unprecedented Best Paper 170+ author list (most didn’t fit on the slide) may be a record for ICRA! So amazing to see what a collaborative community effort can accomplish in pushing robotics + AI forward 🚀

The retina is arguably the most impressive part of the brain. It’s the only part of the brain that faces the world directly - it’s a sensor and processor in one Its consumes 50% more energy per gram than the rest of the brain. 1000:1 compression from retina to optic nerve 25000:1 from optic nerve to brain For every 1Gb collected by the retina, 1Mb is sent to the brain thru the optic nerve and < 100bits used, at a rate of about 875Kbps. Similar to broadband internet.

Don’t miss tomorrow the 6th Robot Learning Workshop at #NeurIPS2023 w/ amazing lineup of speakers: @MashaItkina @_jessethomason_ @DhruvBatraDB @arjunmajum @pathak2206 @SurajNair_1 @BarnesMJ @keerthanpg Montserrat Gonzalez Arenas @xf1280 +robot demos 🤖! robot-learning.ml/2023/

🤖 Excited for the 6th Robot Learning Workshop at #NeurIPS tomorrow! 🚀 Join us for inspiring invited talks, panel, debate, engaging posters, and live robot demos! 🤯 Check out the papers and schedule at robot-learning.ml #RobotLearning #NeurIPS2023 #AI #Robotics

Got a chance to dig through the big robot X-embodiment dataset released last week, and hacked together a little website for others to look through the data. dibyaghosh.com/rtx_viz/ Check it out! There's some pretty random and diverse robot data in there

We have extended the submission deadline of the #NeurIPS2023 #RobotLearning Workshop until Friday, Oct 6. robot-learning.ml @GeorgiaChal @perezjln @HamidrezaKasaei @AlexBewleyAI @shahdhruv_ @m_wulfmeier @RCalandra @igilitschenski @IngmarPosner @DanicaKragic

If you missed the talk then check out the project via human-scene-transformer.github.io

Soon I will be giving a short talk in W04 at #ICCV23 on trajectory prediction for robot navigation on the behalf of @TimSalzmann

Are you interested in cross-embodiment robotics and in need of more data for your models? robotics-transformer-x.github.io

Only a 4 days left to submit to the 6th Robot Learning Workshop: Pretraining, Fine-Tuning, and Generalization with Large Scale Models. robot-learning.ml/2023/

In our Real-time Neural MPC paper, we leverage network capacities 4000x larger in optimizations. We now release L4CasADi, which enables easy integration of PyTorch models in optimizations on CPU and GPU, supporting fast C code generation and seamless integration in Acados.