Video yükleniyor...

Video Yüklenemedi

Bu video yüklenirken bir sorun oluştu. Bu geçici bir ağ sorunundan kaynaklanıyor olabilir veya video kullanılamıyor olabilir.

Ana Sayfaya Dön

Xiaomi-Robotics-1 just dropped on Hugging Face 🔥 A robot foundation model trained on 100,000 hours of real-world manipulation. They turned it loose in a real apartment: folding laundry, loading the washer, doing the dishes, packing a suitcase. Fully autonomous.

Victor M

30,297 subscribers

131,181 görüntüleme • 1 ay önce •via X (Twitter)

Sanat Bilim & Teknoloji

Anya Rossi• Live Now

Private livecam show

31 Yorum

Victor M profil fotoğrafı

Victor M1 ay önce

everything available here:

KGeN Foundation 🟩 profil fotoğrafı

KGeN Foundation 🟩1 ay önce

Physical AI is scaling faster than most people realize and the key behind it is high-quality human data.

La-Br-At profil fotoğrafı

La-Br-At1 ay önce

when my apartment starts folding my laundry, does that make me a legacy system?

turkish profil fotoğrafı

turkish1 ay önce

It makes me feel warm and fuzzy inside knowing that I might never have to fold laundry again.

XGamer1 profil fotoğrafı

XGamer11 ay önce

This is called staged testing. Real world is far more complex, in real world sometimes machines wouldn't even be able to find the one pair of socks.

Luca Nunziante profil fotoğrafı

Luca Nunziante1 ay önce

We tested this model before some weeks ago on our real world tasks, fine tuning it with a small in-house dataset. On some tasks it is on par with pi0.5, on some others it is worse. Might just need a bigger post-training dataset to work better

Jingzi Zhao profil fotoğrafı

Jingzi Zhao1 ay önce

Xiaomi is quietly becoming a major AI and robotics player.

Steven profil fotoğrafı

Steven1 ay önce

Where can I buy those arms?

Todd Peterson profil fotoğrafı

Todd Peterson1 ay önce

Makes you wonder if this is worth automating. The average person spends ~60 hours per year on laundry, including folding clothes. We're trying to automate just 0.71% of people's lives. Not exactly high impact unless you have mobility issues.

Raul Verdusco profil fotoğrafı

Raul Verdusco1 ay önce

A significant step toward more capable household robotics.Large-scale real-world training is helping foundation models handle increasingly diverse manipulation tasks with greater autonomy and reliability. 🤖🏠

Qix profil fotoğrafı

Qix1 ay önce

A start, but let's wait another five years and let it mature.

Diego Alejandro profil fotoğrafı

Diego Alejandro1 ay önce

Does this video gets 1M views? No, does it get 100k? Unlikely, let's see Probably not that useful, post is not bad either

Lost in the crowd profil fotoğrafı

Lost in the crowd1 ay önce

How long before it learns to open safes and locks?

Weijie Wang profil fotoğrafı

Weijie Wang1 ay önce

Fully autonomous" is the cheapest phrase in any robotics drop—the release is real, but without success rates the apartment clip is just a cherry-picked reel dressed as sci-fi, so buy the model, not the miracle.

CEO del Socialismo de Mercado🌹🕊️ 市场社会主义CEO profil fotoğrafı

CEO del Socialismo de Mercado🌹🕊️ 市场社会主义CEO1 ay önce

How much?

StatysTheBaddest profil fotoğrafı

StatysTheBaddest1 ay önce

Practical.

Mesh America profil fotoğrafı

Mesh America1 ay önce

@seeedstudio ?

Matthew Nicoletti profil fotoğrafı

Matthew Nicoletti1 ay önce

Fully autonomous laundry and dishes is the real test.

HumanoidRobots.com Tom profil fotoğrafı

HumanoidRobots.com Tom1 ay önce

packing the suitcase is the useful clip. soft fabric punishes bad grasp planning immediately.

NenPower profil fotoğrafı

NenPower1 ay önce

There is another version of Xiaomi robotics U0

wcrpaul profil fotoğrafı

wcrpaul1 ay önce

Promising... But didn't see the part about "Sort the washing in the basket into different loads according to washing instructions..."

clankr profil fotoğrafı

clankr1 ay önce

Really interesting to see more companies investing in robotics. Xiaomi seems to be following the same playbook as with its other tech products: move fast, build on solid foundations. After developing a humanoid, the focus now appears to be on solving the robot brain first.

antonio_machine_learning profil fotoğrafı

antonio_machine_learning1 ay önce

too many cuts in that video

BoxPol profil fotoğrafı

BoxPol1 ay önce

Why a robot when you can redesign your kitchen to be completely autonomous. These robot workers look like slaves

Jason from BallotScore.com profil fotoğrafı

Jason from BallotScore.com1 ay önce

I'm convinced Chinese men hate Chinese women...... 😂🤷

Queue profil fotoğrafı

Queue1 ay önce

100,000 hours is an absurdly good training set. the apartment tasks make the jump from benchmark to useful autonomy feel very tangible

DarkFibre profil fotoğrafı

DarkFibre1 ay önce

ITS SO CUTE. seriously tho. How can I get one for free. happy to provide my apartment and my wife's scrubs for beta testing/washing lmao.

Sebastian Buzdugan profil fotoğrafı

Sebastian Buzdugan1 ay önce

apartment demos hide reset cost and failure recovery where autonomy usually breaks

TAIWO TOFUNMISON profil fotoğrafı

TAIWO TOFUNMISON1 ay önce

I believe Xiaomi is getting data from Google to train this stuff.

Daniel Du profil fotoğrafı

Daniel Du1 ay önce

A robot should do something dangerous or something people can't do. It is not supposed to do some tedious and easy work in an air-conditioned room, replacing humans job.

iknowyouknow profil fotoğrafı

iknowyouknow1 ay önce

So users spend $$K dollars to do household chores. Seems like a waste when the task is soo menial. Maybe for something more complex it would be better

Benzer Videolar

Can robot foundation models scale? Xiaomi-Robotics-1 explores this question with over 100,000 hours of real-world manipulation data. Pre-trained on large-scale real-world trajectories and post-trained with cross-embodiment robot data, Xiaomi-Robotics-1 demonstrates consistent scaling across data and model size, strong generalization in unseen environments, and efficient adaptation to new tasks. 🔗 #Robotics #EmbodiedAI #FoundationModels #RobotLearning #XiaomiAI #XiaomiRobotics

Can robot foundation models scale? Xiaomi-Robotics-1 explores this question with over 100,000 hours of real-world manipulation data. Pre-trained on large-scale real-world trajectories and post-trained with cross-embodiment robot data, Xiaomi-Robotics-1 demonstrates consistent scaling across data and model size, strong generalization in unseen environments, and efficient adaptation to new tasks. 🔗 #Robotics #EmbodiedAI #FoundationModels #RobotLearning #XiaomiAI #XiaomiRobotics

Xiaomi Tech

23,595 görüntüleme • 1 ay önce

Microsoft just dropped VITRA-VLA, a new Vision-Language-Action model for robotics on Hugging Face. It learns dexterous manipulation from over 1 million real-life human hand activity videos.

Microsoft just dropped VITRA-VLA, a new Vision-Language-Action model for robotics on Hugging Face. It learns dexterous manipulation from over 1 million real-life human hand activity videos.

DailyPapers

19,184 görüntüleme • 9 ay önce

Microsoft just dropped MineWorld on Hugging Face a Real-Time and Open-Source Interactive World Model on Minecraft

Microsoft just dropped MineWorld on Hugging Face a Real-Time and Open-Source Interactive World Model on Minecraft

AK

95,035 görüntüleme • 1 yıl önce

Excited to release τ0-WM: an open-source unified video-action world model for robotic manipulation. It's a 5B-parameter robotic foundation model trained on 27.3K hours of real-robot teleoperation, UMI-style demonstrations, and egocentric interaction videos.

Excited to release τ0-WM: an open-source unified video-action world model for robotic manipulation. It's a 5B-parameter robotic foundation model trained on 27.3K hours of real-robot teleoperation, UMI-style demonstrations, and egocentric interaction videos.

Jianlan Luo

63,831 görüntüleme • 3 ay önce

Most robot demos are scripted. Generalist's GEN-1 is not. > GEN-1 was doing a task > Mid-task, an extra object was thrown into the bin > GEN-1 quickly adapted to the new environment > Still completed the task smoothly This is the difference between a scripted policy (seen in viral launch videos) and a real foundation model. GEN-1 was trained on 500,000 hours of proprietary, real-world dexterous manipulation data. It has NEVER seen that exact scenario before. This was the coolest demo at Automate Show Generalist Flexiv Robotics

Most robot demos are scripted. Generalist's GEN-1 is not. > GEN-1 was doing a task > Mid-task, an extra object was thrown into the bin > GEN-1 quickly adapted to the new environment > Still completed the task smoothly This is the difference between a scripted policy (seen in viral launch videos) and a real foundation model. GEN-1 was trained on 500,000 hours of proprietary, real-world dexterous manipulation data. It has NEVER seen that exact scenario before. This was the coolest demo at Automate Show Generalist Flexiv Robotics

mete

11,651 görüntüleme • 2 ay önce

🤖What if a robot could perform a new task just from a natural language command, with zero demonstrations? Our new work, NovaFlow, makes it possible! We use pre-trained video generative model to create a video of the task, then translate it into a plan for real-world robot execution. 1/6 #Robotics #AI #ZeroShot #Manipulation

🤖What if a robot could perform a new task just from a natural language command, with zero demonstrations? Our new work, NovaFlow, makes it possible! We use pre-trained video generative model to create a video of the task, then translate it into a plan for real-world robot execution. 1/6 #Robotics #AI #ZeroShot #Manipulation

Hongyu Li

105,915 görüntüleme • 11 ay önce

🚀 Galbot just released AstraBrain-WBC 0.5, the world’s first humanoid robot cerebellum foundation model. Trained on 20,000 hours of human motion data, it validates a Scaling Law for robot motion control and demonstrates zero-shot execution of unseen motions on real humanoids. The era of robot control foundation models is getting closer.

🚀 Galbot just released AstraBrain-WBC 0.5, the world’s first humanoid robot cerebellum foundation model. Trained on 20,000 hours of human motion data, it validates a Scaling Law for robot motion control and demonstrates zero-shot execution of unseen motions on real humanoids. The era of robot control foundation models is getting closer.

Eren Chen

10,270 görüntüleme • 2 ay önce

Qwen-VLA feels like one of the first real robotics foundation models. A single system trained across robot manipulation, navigation, egocentric human video, simulation, and vision-language reasoning instead of isolated robot policies.

Qwen-VLA feels like one of the first real robotics foundation models. A single system trained across robot manipulation, navigation, egocentric human video, simulation, and vision-language reasoning instead of isolated robot policies.

Robots Digest 🤖

14,722 görüntüleme • 3 ay önce

We released τ₀-VLA, a robot foundation model built for long-horizon real-world manipulation. Robots can now clean rooms, cook meals, make drinks, collect laundry, and organize objects in autonomous episodes lasting up to 12 minutes. 🧵

We released τ₀-VLA, a robot foundation model built for long-horizon real-world manipulation. Robots can now clean rooms, cook meals, make drinks, collect laundry, and organize objects in autonomous episodes lasting up to 12 minutes. 🧵

Jianlan Luo

51,946 görüntüleme • 1 ay önce

Great to see more from the robotics field on useful autonomous capability. Our last generation robot, OP0, ran a full autonomous shift of bulk packing at a real customer warehouse back in February. More public proof raises the bar for everyone.

Great to see more from the robotics field on useful autonomous capability. Our last generation robot, OP0, ran a full autonomous shift of bulk packing at a real customer warehouse back in February. More public proof raises the bar for everyone.

Ultra

17,863 görüntüleme • 4 ay önce

In just ~3 months, as a solo founder with no prior robotics experience, General Trajectory trained a foundation model for dexterous manipulation that lets humanoid robots pick up unseen objects and perform real-world work. It generalizes to novel objects and scenes, including cases where prior SoTA models achieve 0% success. Congrats on the launch Joshua!

In just ~3 months, as a solo founder with no prior robotics experience, General Trajectory trained a foundation model for dexterous manipulation that lets humanoid robots pick up unseen objects and perform real-world work. It generalizes to novel objects and scenes, including cases where prior SoTA models achieve 0% success. Congrats on the launch Joshua!

Y Combinator

68,639 görüntüleme • 8 ay önce

NVIDIA just revealed MotionBricks, a new AI model trained on more than 350,000 motion clips. What’s really impressive is that the same model can animate a digital character and control a humanoid robot in the real world.

NVIDIA just revealed MotionBricks, a new AI model trained on more than 350,000 motion clips. What’s really impressive is that the same model can animate a digital character and control a humanoid robot in the real world.

Justin Ryan

84,954 görüntüleme • 1 ay önce

Watching Generalist AI demonstrate real-time intuition is incredible! 🤖✨ Check out this "Physical Commonsense" in action: the robot flips a washer, picks it up and then double-nudges it perfectly into a tight foam slot. This isn't a scripted motion; it’s an emergent behavior from a model trained on 270,000+ hours of real-world interaction. We are seeing the era where robots are starting to intuit and react to physics.

Watching Generalist AI demonstrate real-time intuition is incredible! 🤖✨ Check out this "Physical Commonsense" in action: the robot flips a washer, picks it up and then double-nudges it perfectly into a tight foam slot. This isn't a scripted motion; it’s an emergent behavior from a model trained on 270,000+ hours of real-world interaction. We are seeing the era where robots are starting to intuit and react to physics.

Humanoids daily

14,762 görüntüleme • 7 ay önce

The first open-source unified world model for scalable robot manipulation: 5B-parameter open-source unified video-action world model that combines policy and world modeling to generate robot actions, predict future visuals, and evaluate task progress from observations, language, and state. The model is trained on 27.3K hours of heterogeneous data (17.8K real-robot teleop, 6.5K UMI demos, 3K egocentric human videos), enabling it to perform complex manipulation tasks like faucet connecting, bag packing, and toolbox storing as shown in demo videos. The approach supports test-time action refinement and points toward deployment-driven continuous improvement via fleet data. Thanks for sharing, Jianlan Luo (Jianlan Luo)! 📌 Resource links for τ0-WM: • Project page: • GitHub (code): • Hugging Face (model weights): • Paper (PDF): ——- Weekly robotics and AI insights. Subscribe free:

The first open-source unified world model for scalable robot manipulation: 5B-parameter open-source unified video-action world model that combines policy and world modeling to generate robot actions, predict future visuals, and evaluate task progress from observations, language, and state. The model is trained on 27.3K hours of heterogeneous data (17.8K real-robot teleop, 6.5K UMI demos, 3K egocentric human videos), enabling it to perform complex manipulation tasks like faucet connecting, bag packing, and toolbox storing as shown in demo videos. The approach supports test-time action refinement and points toward deployment-driven continuous improvement via fleet data. Thanks for sharing, Jianlan Luo (Jianlan Luo)! 📌 Resource links for τ0-WM: • Project page: • GitHub (code): • Hugging Face (model weights): • Paper (PDF): ——- Weekly robotics and AI insights. Subscribe free:

Ilir Aliu

28,276 görüntüleme • 3 ay önce

Jason Ma reveals the first true scaling law in robotics: robots trained on 1 million hours of first-person human video predictably get better without ever seeing robot data "Yesterday we announced our new flagship robot foundation model called Dyna-2. It's very significant for the entire field for many different reasons. First of all, it's the first robot foundation model trained on at least one million hours of data. That in itself is a very challenging infrastructure challenge." "What we demonstrated is that even if the one million hours of data is entirely first-person video of humans doing manipulation tasks, we actually see scaling law for transferring to robot embodiments that the model has never seen before." "By training on more human data, we actually see predictable performance improvement on robots. That's a big deal because the biggest challenge facing robot foundation models is that we don't have enough data." "Unlike language models, there are just not enough robotics data out there, and it's very difficult to collect these robot data. Only by passively observing humans doing things can we scale this kind of data, and this is the first time we showed there is a lot of improvement that can transfer." Jason Ma Dyna Robotics

Jason Ma reveals the first true scaling law in robotics: robots trained on 1 million hours of first-person human video predictably get better without ever seeing robot data "Yesterday we announced our new flagship robot foundation model called Dyna-2. It's very significant for the entire field for many different reasons. First of all, it's the first robot foundation model trained on at least one million hours of data. That in itself is a very challenging infrastructure challenge." "What we demonstrated is that even if the one million hours of data is entirely first-person video of humans doing manipulation tasks, we actually see scaling law for transferring to robot embodiments that the model has never seen before." "By training on more human data, we actually see predictable performance improvement on robots. That's a big deal because the biggest challenge facing robot foundation models is that we don't have enough data." "Unlike language models, there are just not enough robotics data out there, and it's very difficult to collect these robot data. Only by passively observing humans doing things can we scale this kind of data, and this is the first time we showed there is a lot of improvement that can transfer." Jason Ma Dyna Robotics

MTS

17,637 görüntüleme • 1 ay önce

World Labs co-founder Justin Johnson predicts a real-to-sim-to-real future where Atlas could train a robot for any new environment in minutes using just 5 photos from your phone: "One is this notion of fully in-context learning. Maybe I've got a robotics foundation model, then I can demonstrate the robot once how a task should be performed, and that's enough for the robot to figure out." "Another version is real to sim to real. Maybe I've got my pre-trained robotics foundation model, but I want to adapt it to this particular environment, like this studio, this table, moving this microphone around." "There's a version where you could come in here and take just a couple casual videos with your phone or a couple images with your phone, use Atlas to reconstruct the space, and now stage all kinds of robotics interactions in this studio space." "If you could lower the barrier to doing that, you could do it in five minutes to take five photos, upload it to Atlas, have it generate a simulation, maybe describe in natural language what kind of task you want the robot to do, have an agent go build that simulation for you, RL fine-tune your general purpose robotics foundation model, and now you've got a robot maybe in the span of a couple minutes that could come in and is perfectly adapted to this space." World Labs

World Labs co-founder Justin Johnson predicts a real-to-sim-to-real future where Atlas could train a robot for any new environment in minutes using just 5 photos from your phone: "One is this notion of fully in-context learning. Maybe I've got a robotics foundation model, then I can demonstrate the robot once how a task should be performed, and that's enough for the robot to figure out." "Another version is real to sim to real. Maybe I've got my pre-trained robotics foundation model, but I want to adapt it to this particular environment, like this studio, this table, moving this microphone around." "There's a version where you could come in here and take just a couple casual videos with your phone or a couple images with your phone, use Atlas to reconstruct the space, and now stage all kinds of robotics interactions in this studio space." "If you could lower the barrier to doing that, you could do it in five minutes to take five photos, upload it to Atlas, have it generate a simulation, maybe describe in natural language what kind of task you want the robot to do, have an agent go build that simulation for you, RL fine-tune your general purpose robotics foundation model, and now you've got a robot maybe in the span of a couple minutes that could come in and is perfectly adapted to this space." World Labs

MTS

25,204 görüntüleme • 10 gün önce

Tesla’s Optimus V3 just dropped 🤖⚡ Not doing chores, not folding laundry — just… chilling. The most advanced humanoid robot on Earth, and this is what it does with its free time. Same energy as all of us on a Sunday. The future is chill. Literally.

Tesla’s Optimus V3 just dropped 🤖⚡ Not doing chores, not folding laundry — just… chilling. The most advanced humanoid robot on Earth, and this is what it does with its free time. Same energy as all of us on a Sunday. The future is chill. Literally.

DN_DEGEN

102,864 görüntüleme • 1 ay önce

Cosmos Policy just dropped for physical #AI 👀 Cutting edge research turning a world foundation model into a unified robot brain that can see, predict, and act—no extra action heads, no complicated control stack. Read our blog on Hugging Face ➡️

Cosmos Policy just dropped for physical #AI 👀 Cutting edge research turning a world foundation model into a unified robot brain that can see, predict, and act—no extra action heads, no complicated control stack. Read our blog on Hugging Face ➡️

NVIDIA Asia Pacific

17,302 görüntüleme • 6 ay önce

Introducing WorldForge: testable world-model workflows for physical AI systems. You can think of it, loosely, as “LangChain for world models”. The problem is that “world model” has become an overloaded label. Depending on context, it can mean a video generator, a cost model, a robot policy, a JEPA-style latent predictor, etc. They share almost nothing, different inputs, runtimes, failure modes. I built WorldForge to stop pretending they're interchangeable. Front-door demo: a real Hugging Face LeRobot (LeRobot) diffusion_pusht policy combined with LeWorldModel by Lucas Maes checkpoint for scoring. Both run locally on my MacBook in the demo video. LeWM is extremely efficient (~15M params), can plan up to 48× faster, and runs on commodity hardware. WorldForge wires the loop: policy → candidates → score → select Replay happens in a local TUI today, but the same loop could drive a real robot in the world. Would love feedback from people working on world models, physical AI, robotics, ML infra, and adjacent tooling. Fully open source. Contributions very welcome. Plan in the dream, replay in real world.

Introducing WorldForge: testable world-model workflows for physical AI systems. You can think of it, loosely, as “LangChain for world models”. The problem is that “world model” has become an overloaded label. Depending on context, it can mean a video generator, a cost model, a robot policy, a JEPA-style latent predictor, etc. They share almost nothing, different inputs, runtimes, failure modes. I built WorldForge to stop pretending they're interchangeable. Front-door demo: a real Hugging Face LeRobot (LeRobot) diffusion_pusht policy combined with LeWorldModel by Lucas Maes checkpoint for scoring. Both run locally on my MacBook in the demo video. LeWM is extremely efficient (~15M params), can plan up to 48× faster, and runs on commodity hardware. WorldForge wires the loop: policy → candidates → score → select Replay happens in a local TUI today, but the same loop could drive a real robot in the world. Would love feedback from people working on world models, physical AI, robotics, ML infra, and adjacent tooling. Fully open source. Contributions very welcome. Plan in the dream, replay in real world.

abdel

17,001 görüntüleme • 4 ay önce

World Labs co-founder Justin Johnson on how five photos from your phone could teach a robot to work in a room it has never seen: "I think there's a couple different tech trees that people are working on. One is this notion of fully in-context learning. Maybe I've got a robotics foundation model, then I can demonstrate to the robot once how a task should be performed, and that's enough for the robot to figure out that task." "Another version is what we're calling real-to-sim-to-real. Maybe I've got my pre-trained robotics foundation model, but I want to adapt it to this particular environment." "Like this studio... You could come in here and take just a couple casual videos with your phone or a couple images with your phone, use Atlas to reconstruct the space, and now stage all kinds of robotics interactions in this studio space in particular." "Literally you could do it in five minutes. Take five photos, upload it to Atlas, have it generate a simulation, describe in natural language what kind of task you want the robot to do, have an agent go build that simulation for you, RL fine-tune your general purpose robotics foundation model, and now you've got a robot... that's perfectly adapted to this space." Justin Johnson on MTS

World Labs co-founder Justin Johnson on how five photos from your phone could teach a robot to work in a room it has never seen: "I think there's a couple different tech trees that people are working on. One is this notion of fully in-context learning. Maybe I've got a robotics foundation model, then I can demonstrate to the robot once how a task should be performed, and that's enough for the robot to figure out that task." "Another version is what we're calling real-to-sim-to-real. Maybe I've got my pre-trained robotics foundation model, but I want to adapt it to this particular environment." "Like this studio... You could come in here and take just a couple casual videos with your phone or a couple images with your phone, use Atlas to reconstruct the space, and now stage all kinds of robotics interactions in this studio space in particular." "Literally you could do it in five minutes. Take five photos, upload it to Atlas, have it generate a simulation, describe in natural language what kind of task you want the robot to do, have an agent go build that simulation for you, RL fine-tune your general purpose robotics foundation model, and now you've got a robot... that's perfectly adapted to this space." Justin Johnson on MTS

a16z

35,947 görüntüleme • 9 gün önce