Loading video...

Video Failed to Load

Go Home

Robot Utility Models (RUMs) enable basic tasks โ€“ door opening, drawer opening, object reorientation, etc. โ€“ at ~90% accuracy without ANY finetuning (i.e. zero-shot) in unseen new environments. Fully open source!!! models, data, code & hw. We think this is super exciting, why?๐Ÿ‘‡ 1. Unlocks many practical home utility...

notmahi's profile picture

Mahi Shafiullah ๐Ÿ ๐Ÿค–

3,074 subscribers

89,535 views โ€ข 1 year ago โ€ขvia X (Twitter)

Science & TechnologyHealth & WellnessEducation
LIVE

Anya Rossiโ€ข Live Now

Private livecam show

0 Comments

No comments available

Comments from the original post will appear here

Related Videos

Predicting how the environment changes over time is a key task in robotics, enabling all kinds of downstream tasks. Trace Anything is a very interesting look at how you can train models which predict the trajectory each point will take in the future -- you can use this for robot manipulation, video prediction, and more. Learn more ->
0:16

Predicting how the environment changes over time is a key task in robotics, enabling all kinds of downstream tasks. Trace Anything is a very interesting look at how you can train models which predict the trajectory each point will take in the future -- you can use this for robot manipulation, video prediction, and more. Learn more ->

Chris Paxton

56,845 views โ€ข 5 months ago

How can we more effectively leverage robot data from different embodiments for skill transfer? Excited to share that our new work, RoVi-Aug, has been accepted to Conference on Robot Learning as an oral paper๏ผ WIth RoVi-Aug, you can augment an existing robot dataset into a different robot and different viewpoints. A policy trained on the augmented dataset can zero-shot deploy on the unseen target robot with significantly different camera angles! ๐Ÿงต๐Ÿ‘‡ ๐Ÿ”— Check out our paper:
0:18

How can we more effectively leverage robot data from different embodiments for skill transfer? Excited to share that our new work, RoVi-Aug, has been accepted to Conference on Robot Learning as an oral paper๏ผ WIth RoVi-Aug, you can augment an existing robot dataset into a different robot and different viewpoints. A policy trained on the augmented dataset can zero-shot deploy on the unseen target robot with significantly different camera angles! ๐Ÿงต๐Ÿ‘‡ ๐Ÿ”— Check out our paper:

Chenfeng_X

28,278 views โ€ข 1 year ago

August came and went in a flash and what a month it was! We opened 15 new stores (including the opening of our first store in Zimbabwe ๐Ÿ‡ฟ๐Ÿ‡ผ) and completed 1 revamp. Things are really ramping up as we head into the final stretch of the year, with more new countries and many more store openings still in the pipeline for 2025. ๐Ÿš€ The momentum is building, and weโ€™re just getting started!
0:15

August came and went in a flash and what a month it was! We opened 15 new stores (including the opening of our first store in Zimbabwe ๐Ÿ‡ฟ๐Ÿ‡ผ) and completed 1 revamp. Things are really ramping up as we head into the final stretch of the year, with more new countries and many more store openings still in the pipeline for 2025. ๐Ÿš€ The momentum is building, and weโ€™re just getting started!

Hungry Lion - SA

259,207 views โ€ข 9 months ago

World Models are the path for some AI Models in the future. But how can we efficiently train these models to not only see the world the way humans do but to see the world in a new and unique way. By visualizing, what is normally sequenced audio patterns, we can derive much more insights. Here we see Paganini in a visual form that can than be described and transcribed into a World Model. We can observe connections in a manner that may not have been clear prior to the digitalization of music and sound in this way. The company with the most valuable potential in building a World Model is Tesla. Not that this type of visualization is being used, but that the mechanisms are in place, and the technology is in place for the company to thrive in this new form of AI.
0:19

World Models are the path for some AI Models in the future. But how can we efficiently train these models to not only see the world the way humans do but to see the world in a new and unique way. By visualizing, what is normally sequenced audio patterns, we can derive much more insights. Here we see Paganini in a visual form that can than be described and transcribed into a World Model. We can observe connections in a manner that may not have been clear prior to the digitalization of music and sound in this way. The company with the most valuable potential in building a World Model is Tesla. Not that this type of visualization is being used, but that the mechanisms are in place, and the technology is in place for the company to thrive in this new form of AI.

Brian Roemmele

57,424 views โ€ข 6 months ago

Train a TensorFlow object detection model โ€“ then deploy it on a robot ๐Ÿค– Iulia Feroli (Iulia Feroli) shows how to turn a notebook into a real-time object detection app. This tutorial works for any project โ€“ though we demonstrate the deployment on (and assisted by!) the #ReachyMini, an open-source robot from Pollen Robotics. Built with PyCharm + Claude Code. ๐Ÿ‘‰ Watch it in action:
0:17

Train a TensorFlow object detection model โ€“ then deploy it on a robot ๐Ÿค– Iulia Feroli (Iulia Feroli) shows how to turn a notebook into a real-time object detection app. This tutorial works for any project โ€“ though we demonstrate the deployment on (and assisted by!) the #ReachyMini, an open-source robot from Pollen Robotics. Built with PyCharm + Claude Code. ๐Ÿ‘‰ Watch it in action:

PyCharm, a JetBrains IDE

34,908 views โ€ข 1 month ago

Proud to announce DobbยทE: the next step in home robot system that I was working on for the past 3 years. We have visited 10 homes, learned 100+ tasks, and we are just getting started! And we fully open-sourced it all, hardware, models, and software: ๐Ÿงต
0:17

Proud to announce DobbยทE: the next step in home robot system that I was working on for the past 3 years. We have visited 10 homes, learned 100+ tasks, and we are just getting started! And we fully open-sourced it all, hardware, models, and software: ๐Ÿงต

Mahi Shafiullah ๐Ÿ ๐Ÿค–

164,452 views โ€ข 2 years ago

Opening doors to Headout's new HQ in BLR. A space we imagined as a Museum of the World, designed as an ode to the big, beautiful world around us. We made it with love, and thereโ€™s room for many more. Weโ€™re hiring across roles. Come build with us.
0:10

Opening doors to Headout's new HQ in BLR. A space we imagined as a Museum of the World, designed as an ode to the big, beautiful world around us. We made it with love, and thereโ€™s room for many more. Weโ€™re hiring across roles. Come build with us.

Varun Khona

39,392 views โ€ข 1 month ago

My favorite paper this year: "Video models are zero-shot learners and reasoners" It illustrates that video models show emergent visual reasoning at scale - they can solve vision tasks they weren't trained for. This may be the "GPT moment" for vision. Let's break it down ๐Ÿ‘‡
0:19

My favorite paper this year: "Video models are zero-shot learners and reasoners" It illustrates that video models show emergent visual reasoning at scale - they can solve vision tasks they weren't trained for. This may be the "GPT moment" for vision. Let's break it down ๐Ÿ‘‡

Justine Moore

185,001 views โ€ข 5 months ago

๐Ÿค– Another zero-shot reward model is now in LeRobot: ROBOMETER. A general-purpose, zero-shot video-language reward model from University of South Carolina, UT Dallas, Massachusetts Institute of Technology (MIT), University of Washington, Ai2, and NVIDIA that predicts frame-level task progress. Trained on 1M+ trajectories from 21 robot embodiments, generalizes zero-shot to unseen tasks, scenes, and robots. 2.4โ€“4.5x better downstream success rates across online RL, offline RL, data filtering, failure detection, and data retrieval for IL. Project: Paper:
0:14

๐Ÿค– Another zero-shot reward model is now in LeRobot: ROBOMETER. A general-purpose, zero-shot video-language reward model from University of South Carolina, UT Dallas, Massachusetts Institute of Technology (MIT), University of Washington, Ai2, and NVIDIA that predicts frame-level task progress. Trained on 1M+ trajectories from 21 robot embodiments, generalizes zero-shot to unseen tasks, scenes, and robots. 2.4โ€“4.5x better downstream success rates across online RL, offline RL, data filtering, failure detection, and data retrieval for IL. Project: Paper:

LeRobot

31,605 views โ€ข 8 days ago

The opening of NODE feels special. A real home for the CryptoPunks and a meaningful step for NFTs as digital art. Less noise, more culture. This is about understanding, preserving, and taking digital art seriously. A new chapter is starting.
2:27

The opening of NODE feels special. A real home for the CryptoPunks and a meaningful step for NFTs as digital art. Less noise, more culture. This is about understanding, preserving, and taking digital art seriously. A new chapter is starting.

THฮžGABO๐ŸŒ

12,315 views โ€ข 4 months ago

Introducing $SAHARA โ€” the native utility token of Sahara AI. Designed for real-world function, $SAHARA powers a new economic model for AI. One that is collaborative, open, and built to reward real contributors. Utility, distribution, and claim details below ๐Ÿงต๐Ÿ‘‡
0:17

Introducing $SAHARA โ€” the native utility token of Sahara AI. Designed for real-world function, $SAHARA powers a new economic model for AI. One that is collaborative, open, and built to reward real contributors. Utility, distribution, and claim details below ๐Ÿงต๐Ÿ‘‡

Sahara AI ๐Ÿ”†

845,184 views โ€ข 11 months ago

Robotics keeps hitting the same wall. Single task RL works, but... it does not scale to hundreds of tasks or new embodiments. This new paper looks like a real step toward fixing that. The team introduces MMBench, a benchmark with 200 tasks across many domains and robots, and Newt, a language conditioned world model trained online across all 200 tasks at once. The simple idea behind Newt: The model learns from demos to get the right priors It trains across many tasks through online interaction It uses language to ground the goal It adapts fast when a new task shows up What stood out to me: โœ… One model trained on 200 tasks at the same time โœ… Language conditioned control for both states and RGB โœ… Better data efficiency than strong baselines โœ… Strong open loop control โœ… Fast adaptation to new tasks and embodiments โœ… Full release of 200 checkpoints, 4000 demos, code, and benchmark This is a good push toward general control instead of one model per task. If you want the full paper: Project page: โ€”- Weekly robotics and AI insights. Subscribe free:
0:11

Robotics keeps hitting the same wall. Single task RL works, but... it does not scale to hundreds of tasks or new embodiments. This new paper looks like a real step toward fixing that. The team introduces MMBench, a benchmark with 200 tasks across many domains and robots, and Newt, a language conditioned world model trained online across all 200 tasks at once. The simple idea behind Newt: The model learns from demos to get the right priors It trains across many tasks through online interaction It uses language to ground the goal It adapts fast when a new task shows up What stood out to me: โœ… One model trained on 200 tasks at the same time โœ… Language conditioned control for both states and RGB โœ… Better data efficiency than strong baselines โœ… Strong open loop control โœ… Fast adaptation to new tasks and embodiments โœ… Full release of 200 checkpoints, 4000 demos, code, and benchmark This is a good push toward general control instead of one model per task. If you want the full paper: Project page: โ€”- Weekly robotics and AI insights. Subscribe free:

Ilir Aliu

70,090 views โ€ข 6 months ago

After 5 months of planning, our first Live2D Jakarta Chapter event was a huge success! We brought an interactive Live2D showcase to the biggest anime con in Indonesia, featuring over 20 finest models created by our local artists and riggers. This showcase also included a custom-made plugin for VTube Studio that adds touch interactivity to Live2D models, allowing visitors to interact with VTuber models in a more engaging way! We hope this demo could open up new potential for Vtubers in IRL events. Special thanks to Live2D Inc. for trusting me to host this event! The showcase attracted many curious visitors and gave us the opportunity to introduce this tech to a wider audience, and we hope we are opening up new opportunities for everyone. Look forward to our next one! (Model: Key Oriesa by karamomo๐Ÿ‘ | Live2D Rigging | Food Illustration)
0:19

After 5 months of planning, our first Live2D Jakarta Chapter event was a huge success! We brought an interactive Live2D showcase to the biggest anime con in Indonesia, featuring over 20 finest models created by our local artists and riggers. This showcase also included a custom-made plugin for VTube Studio that adds touch interactivity to Live2D models, allowing visitors to interact with VTuber models in a more engaging way! We hope this demo could open up new potential for Vtubers in IRL events. Special thanks to Live2D Inc. for trusting me to host this event! The showcase attracted many curious visitors and gave us the opportunity to introduce this tech to a wider audience, and we hope we are opening up new opportunities for everyone. Look forward to our next one! (Model: Key Oriesa by karamomo๐Ÿ‘ | Live2D Rigging | Food Illustration)

Ran ๐Ÿฒ Live2D Animator

163,183 views โ€ข 22 days ago

What would it look like to see all ~13,000 proteins inside a single human cell at once? We share ProtiCelli โ€” a generative model that makes this possible. Besides benchmarking, we demonstrate utility in many tasks. Preprint: A thread ๐Ÿงต๐Ÿ‘‡
0:16

What would it look like to see all ~13,000 proteins inside a single human cell at once? We share ProtiCelli โ€” a generative model that makes this possible. Besides benchmarking, we demonstrate utility in many tasks. Preprint: A thread ๐Ÿงต๐Ÿ‘‡

Emma Lundberg

37,695 views โ€ข 2 months ago

SuperDapp is LIVE! ๐Ÿค– The future of AI-powered Web3 interaction starts NOW. A new era of intelligent AI agents, SocialFi, and seamless blockchain integration is here. This is more than just an appโ€”it's a movement. Today we launch our major updates, new modes, dapp improvements, and much more. ๐ŸŒ Welcome to the future of web3 utility. Today will be an exciting day of announcements ๐Ÿ”ฅ #SuperDapp ๐Ÿงต๐Ÿ‘‡
0:10

SuperDapp is LIVE! ๐Ÿค– The future of AI-powered Web3 interaction starts NOW. A new era of intelligent AI agents, SocialFi, and seamless blockchain integration is here. This is more than just an appโ€”it's a movement. Today we launch our major updates, new modes, dapp improvements, and much more. ๐ŸŒ Welcome to the future of web3 utility. Today will be an exciting day of announcements ๐Ÿ”ฅ #SuperDapp ๐Ÿงต๐Ÿ‘‡

SuperDapp

61,254 views โ€ข 1 year ago

CCTV has revealed the production process of a new type of missile for the first time, and it is truly eye-opening. This fully automated, unmanned, intelligent weapon has received a surge in orders and is a highly coveted product for many countries.#China
0:17

CCTV has revealed the production process of a new type of missile for the first time, and it is truly eye-opening. This fully automated, unmanned, intelligent weapon has received a surge in orders and is a highly coveted product for many countries.#China

Johannes Maria

11,598 views โ€ข 1 year ago

I tried the knocking shot with Kling 2.1, and we still have a failure clearly the man is not touching the door at all. Anyone have access to Veo3 please try it and let me know if this works! it seems like a simple shot, but all models failed to do it, let's see if Veo3 can๐Ÿ”ฅ
0:10

I tried the knocking shot with Kling 2.1, and we still have a failure clearly the man is not touching the door at all. Anyone have access to Veo3 please try it and let me know if this works! it seems like a simple shot, but all models failed to do it, let's see if Veo3 can๐Ÿ”ฅ

Amira Zairi

14,527 views โ€ข 1 year ago

โ€œIโ€™m here for the future โ€” when everything is onchain, and we move beyond Bitcoin and memecoins into a place where crypto is actually very useful, and it confers fundamental utility to participants. I think thatโ€™s going to be exciting โ€” and the time has come. I think itโ€™s this year.โ€ โ€” Vlad Tenev, on Robinhood Chain, an ethereum L2
0:18

โ€œIโ€™m here for the future โ€” when everything is onchain, and we move beyond Bitcoin and memecoins into a place where crypto is actually very useful, and it confers fundamental utility to participants. I think thatโ€™s going to be exciting โ€” and the time has come. I think itโ€™s this year.โ€ โ€” Vlad Tenev, on Robinhood Chain, an ethereum L2

rip.eth

20,541 views โ€ข 11 months ago