Loading video...

Video Failed to Load

There was a problem loading this video. This could be due to a temporary network issue or the video might be unavailable.

Introduce EgoInfinity: a web-scale (14.6 yrs, 142M clips) data engine that automatically lifts Youtube videos into 4D hand-object interaction, and retargets for robot learning Not only a dataset, but a modular, upgradable engine HF Space:

Vector Wang

2,225 subscribers

107,906 views • 1 month ago •via X (Twitter)

Education Science & Technology

Anya Rossi• Live Now

Private livecam show

0 Comments

No comments available

Comments from the original post will appear here

Related Videos

HDMI (HumanoiD iMitation for Interaction) is a framework enabling humanoid robots to learn whole-body object interaction skills from monocular RGB human videos. It extracts and retargets human poses and object trajectories using GVHMR and LocoMujoco, building reference datasets with contact annotations. The data is used to train an RL policy via robot-object co-tracking. HDMI achieved 67 consecutive door traversals.

HDMI (HumanoiD iMitation for Interaction) is a framework enabling humanoid robots to learn whole-body object interaction skills from monocular RGB human videos. It extracts and retargets human poses and object trajectories using GVHMR and LocoMujoco, building reference datasets with contact annotations. The data is used to train an RL policy via robot-object co-tracking. HDMI achieved 67 consecutive door traversals.

The Humanoid Hub

17,395 views • 10 months ago

Toyota IMV Origin is not just a pickup truck, but a modular work platform that the company intentionally makes "unfinished." Toyota supplies only the base — the body, chassis, and engine. Everything else is assembled by the owners.

Toyota IMV Origin is not just a pickup truck, but a modular work platform that the company intentionally makes "unfinished." Toyota supplies only the base — the body, chassis, and engine. Everything else is assembled by the owners.

Massimo

193,722 views • 9 months ago

HumanX: a scalable framework that converts single monocular human videos into agile, generalizable interaction skills for humanoid robots without task-specific rewards. Core parts: - XGen: retargets human motion + synthesizes diverse physically plausible training data (via physics-driven object trajectories and augmentation). - XMimic: unified imitation pipeline for robust generalization. Shows promising results for both Blind (no external sensors) and MoCap modes.

HumanX: a scalable framework that converts single monocular human videos into agile, generalizable interaction skills for humanoid robots without task-specific rewards. Core parts: - XGen: retargets human motion + synthesizes diverse physically plausible training data (via physics-driven object trajectories and augmentation). - XMimic: unified imitation pipeline for robust generalization. Shows promising results for both Blind (no external sensors) and MoCap modes.

The Humanoid Hub

20,358 views • 6 months ago

Perhaps the "GPT-3 moment" for humanoid robots lies within vast amounts of human video data. UC Berkeley (Abbeel + Malik group) just released Do as I Do: a pipeline that takes monocular RGB human videos (ego + exo)->reconstructs hand-object interactions->retargets them to multi-fingered robotic hands, generating executable robot trajectories. It pushes success rates on real-world clips from ~25% to 71%, outperforming prior SOTA. This kind of human video->robot-complete data approach feels increasingly important for closing the data flywheel in humanoid/embodied AI.

Perhaps the "GPT-3 moment" for humanoid robots lies within vast amounts of human video data. UC Berkeley (Abbeel + Malik group) just released Do as I Do: a pipeline that takes monocular RGB human videos (ego + exo)->reconstructs hand-object interactions->retargets them to multi-fingered robotic hands, generating executable robot trajectories. It pushes success rates on real-world clips from ~25% to 71%, outperforming prior SOTA. This kind of human video->robot-complete data approach feels increasingly important for closing the data flywheel in humanoid/embodied AI.

CyberRobo

12,107 views • 1 month ago

ViPE: Video Pose Engine for 3D Geometric Perception Contributions: • A robust and efficient framework, ViPE, for estimating camera parameters and dense depth from diverse, in-the-wild videos. • A system design that integrates the strengths of classical SLAM (efficiency, scalability) and learned models (robustness), with key improvements in efficiency, dynamic object handling, and depth quality over prior work. • A large-scale dataset of annotated videos, created using ViPE, to facilitate future research in 3D computer vision.

ViPE: Video Pose Engine for 3D Geometric Perception Contributions: • A robust and efficient framework, ViPE, for estimating camera parameters and dense depth from diverse, in-the-wild videos. • A system design that integrates the strengths of classical SLAM (efficiency, scalability) and learned models (robustness), with key improvements in efficiency, dynamic object handling, and depth quality over prior work. • A large-scale dataset of annotated videos, created using ViPE, to facilitate future research in 3D computer vision.

MrNeRF

42,553 views • 1 year ago

Want a robot that learns household tasks by watching you? EquiBot is a ✨ generalizable and 🚰 data-efficient method for visuomotor policy learning, robust to changes in object shapes, lighting, and scene makeup, even from just 5 mins of human videos. 🧵↓

Want a robot that learns household tasks by watching you? EquiBot is a ✨ generalizable and 🚰 data-efficient method for visuomotor policy learning, robust to changes in object shapes, lighting, and scene makeup, even from just 5 mins of human videos. 🧵↓

Jingyun Yang

88,399 views • 2 years ago

Large-scale 3D Scene Generation (all scenes are real-time rendered)!! Physically-grounded generative data without hallucinations is the missing link for robot learning and testing at scale. We introduce a method that directly generates large-scale 3D driving scenes with accurate geometry, allowing for causal view synthesis and generation with object permanence and explicit 3D geometry. This also allows for extreme trajectory extrapolation without failure! We also show that we can build fully data-driven simulators for end-to-end learning with this approach. Project: with the amazing team of Julian Ost, Amogh Joshi , Andrea Ramazzina, Maximilian Bömer, Mario Bijelic.

Large-scale 3D Scene Generation (all scenes are real-time rendered)!! Physically-grounded generative data without hallucinations is the missing link for robot learning and testing at scale. We introduce a method that directly generates large-scale 3D driving scenes with accurate geometry, allowing for causal view synthesis and generation with object permanence and explicit 3D geometry. This also allows for extreme trajectory extrapolation without failure! We also show that we can build fully data-driven simulators for end-to-end learning with this approach. Project: with the amazing team of Julian Ost, Amogh Joshi , Andrea Ramazzina, Maximilian Bömer, Mario Bijelic.

Felix Heide

27,779 views • 11 months ago

The World Model as NEO's Cognitive Core 1X has revealed a major AI development where the NEO humanoid can translate any natural language prompt into robotic action. It demonstrates this capability even for novel tasks, objects, and environments not found in its robot dataset. - the 1X World Model is trained on internet-scale human interaction videos and fine-tuned with robot data to ground its understanding in physics and in NEO's embodiment - from a simple voice or text prompt, the world model generates a visualization of future actions - a built-in inverse dynamics model then translates these into precise motor movements for NEO

The World Model as NEO's Cognitive Core 1X has revealed a major AI development where the NEO humanoid can translate any natural language prompt into robotic action. It demonstrates this capability even for novel tasks, objects, and environments not found in its robot dataset. - the 1X World Model is trained on internet-scale human interaction videos and fine-tuned with robot data to ground its understanding in physics and in NEO's embodiment - from a simple voice or text prompt, the world model generates a visualization of future actions - a built-in inverse dynamics model then translates these into precise motor movements for NEO

The Humanoid Hub

68,453 views • 7 months ago

I don’t know if we live in a Matrix, but I know for sure that robots will spend most of their lives in simulation. Let machines train machines. I’m excited to introduce DexMimicGen, a massive-scale synthetic data generator that enables a humanoid robot to learn complex skills from only a handful of human demonstrations. Yes, as few as 5! DexMimicGen addresses the biggest pain point in robotics: where do we get data? Unlike with LLMs, where vast amounts of texts are readily available, you cannot simply download motor control signals from the internet. So researchers teleoperate the robots to collect motion data via XR headsets. They have to repeat the same skill over and over and over again, because neural nets are data hungry. This is a very slow and uncomfortable process. At NVIDIA, we believe the majority of high-quality tokens for robot foundation models will come from simulation. What DexMimicGen does is to trade GPU compute time for human time. It takes one motion trajectory from human, and multiplies into 1000s of new trajectories. A robot brain trained on this augmented dataset will generalize far better in the real world. Think of DexMimicGen as a learning signal amplifier. It maps a small dataset to a large (de facto infinite) dataset, using physics simulation in the loop. In this way, we free humans from babysitting the bots all day. The future of robot data is generative. The future of the entire robot learning pipeline will also be generative. 🧵

I don’t know if we live in a Matrix, but I know for sure that robots will spend most of their lives in simulation. Let machines train machines. I’m excited to introduce DexMimicGen, a massive-scale synthetic data generator that enables a humanoid robot to learn complex skills from only a handful of human demonstrations. Yes, as few as 5! DexMimicGen addresses the biggest pain point in robotics: where do we get data? Unlike with LLMs, where vast amounts of texts are readily available, you cannot simply download motor control signals from the internet. So researchers teleoperate the robots to collect motion data via XR headsets. They have to repeat the same skill over and over and over again, because neural nets are data hungry. This is a very slow and uncomfortable process. At NVIDIA, we believe the majority of high-quality tokens for robot foundation models will come from simulation. What DexMimicGen does is to trade GPU compute time for human time. It takes one motion trajectory from human, and multiplies into 1000s of new trajectories. A robot brain trained on this augmented dataset will generalize far better in the real world. Think of DexMimicGen as a learning signal amplifier. It maps a small dataset to a large (de facto infinite) dataset, using physics simulation in the loop. In this way, we free humans from babysitting the bots all day. The future of robot data is generative. The future of the entire robot learning pipeline will also be generative. 🧵

Jim Fan

165,246 views • 1 year ago

🚀 Struggling with the lack of high-quality data for AI-driven human-object interaction research? We've got you covered! Introducing HUMOTO, a groundbreaking 4D dataset for human-object interaction, developed with a combination of wearable motion capture, SOTA 6D pose estimation vision models, LLM, and the professional refining works of multiple animation studios. HUMOTO features: ✅ Over 700 diverse daily activities ✅ Interactions with 60+ objects, 70+ articulated parts. ✅ Fine-grained text annotations ✅ Detailed hand and finger movements We hope HUMOTO will fuel your Humanoid AI research and drive new advancements! For research or commercial license inquiries, please contact yizho@adobe.com. Explore the dataset: 👉 HUMOTO Dataset Website Learn more: 👉 HUMOTO Project Page Jiaxin Lu Qixing Huang #AI #HumanObjectInteraction #HumanoidAI #MotionCapture #HUMOTO #AdobeResearch

🚀 Struggling with the lack of high-quality data for AI-driven human-object interaction research? We've got you covered! Introducing HUMOTO, a groundbreaking 4D dataset for human-object interaction, developed with a combination of wearable motion capture, SOTA 6D pose estimation vision models, LLM, and the professional refining works of multiple animation studios. HUMOTO features: ✅ Over 700 diverse daily activities ✅ Interactions with 60+ objects, 70+ articulated parts. ✅ Fine-grained text annotations ✅ Detailed hand and finger movements We hope HUMOTO will fuel your Humanoid AI research and drive new advancements! For research or commercial license inquiries, please contact [email protected]. Explore the dataset: 👉 HUMOTO Dataset Website Learn more: 👉 HUMOTO Project Page Jiaxin Lu Qixing Huang #AI #HumanObjectInteraction #HumanoidAI #MotionCapture #HUMOTO #AdobeResearch

Yi Zhou

71,609 views • 1 year ago

What we’re building with MasterBOT goes far beyond a single robot It’s a framework for training autonomous intelligence at scale A decentralized simulation engine that accelerates how machines learn, adapt, and evolve Tech giants are building the future of robotics behind closed doors We're building it in the open The $BOT revolution has only just begun

What we’re building with MasterBOT goes far beyond a single robot It’s a framework for training autonomous intelligence at scale A decentralized simulation engine that accelerates how machines learn, adapt, and evolve Tech giants are building the future of robotics behind closed doors We're building it in the open The $BOT revolution has only just begun

That Investor

167,649 views • 10 months ago

Over the last few months, we’ve been thinking about how to learn from “off-domain” data - data from non-robot sources like video or simulation. These data sources are not quite good enough to learn policies (even monolithic VLA models) directly, but they still contain lots of information that can be useful for generalizable robot control. How can we develop robot learning models that are able to make use of this type of data for generalizable control? In new work, that we call HAMSTER, we show that VLMs can be useful for enabling robotic learning from off-domain data, but specifically when used through hierarchical VLA architectures. We show that this class of models can learn generalizable robot policies for the real world from large-scale, off-domain data. A 🧵 (1/10)

Over the last few months, we’ve been thinking about how to learn from “off-domain” data - data from non-robot sources like video or simulation. These data sources are not quite good enough to learn policies (even monolithic VLA models) directly, but they still contain lots of information that can be useful for generalizable robot control. How can we develop robot learning models that are able to make use of this type of data for generalizable control? In new work, that we call HAMSTER, we show that VLMs can be useful for enabling robotic learning from off-domain data, but specifically when used through hierarchical VLA architectures. We show that this class of models can learn generalizable robot policies for the real world from large-scale, off-domain data. A 🧵 (1/10)

Abhishek Gupta

11,994 views • 1 year ago

So I heard we need more data for robot learning :) Purely real world teleop is expensive and slow, making large scale data collection challenging. I’ve been excited about getting more data into robot learning, going beyond just real-world teleop data. To this end, we’ve been scaling up data generation with RL in realistic simulations generated on the fly from crowdsourced videos. Enables realistic data collection, much more cheaply than purely real world teleop. Importantly, data collection becomes even*cheaper* with more environments, allowing training with over 100x more data. Transfers to real robots for generalizable manipulation. A 🧵 (1/N)

So I heard we need more data for robot learning :) Purely real world teleop is expensive and slow, making large scale data collection challenging. I’ve been excited about getting more data into robot learning, going beyond just real-world teleop data. To this end, we’ve been scaling up data generation with RL in realistic simulations generated on the fly from crowdsourced videos. Enables realistic data collection, much more cheaply than purely real world teleop. Importantly, data collection becomes evencheaper with more environments, allowing training with over 100x more data. Transfers to real robots for generalizable manipulation. A 🧵 (1/N)

Abhishek Gupta

13,350 views • 1 year ago

Microsoft has launched a powerful new data analysis tool! Introducing Data Formulator, a 100% open-source LLM-powered, no-code tool that transforms data in a snap and creates stunning visualizations. Key features include: 🤖 AI-powered data transformation 🖱️ Interactive drag-and-drop UI for visualizations 💬 Seamless blend of UI & natural language inputs But that’s not all: You can even create charts beyond your initial dataset. Data Formulator automatically identifies extra computation needs, generates fields for you, and outputs the final visualization. Find the GitHub repo in the next tweet! _____ Find me → Akshay 🚀 ✔️ For more insights and tutorials on AI and Machine Learning.

Microsoft has launched a powerful new data analysis tool! Introducing Data Formulator, a 100% open-source LLM-powered, no-code tool that transforms data in a snap and creates stunning visualizations. Key features include: 🤖 AI-powered data transformation 🖱️ Interactive drag-and-drop UI for visualizations 💬 Seamless blend of UI & natural language inputs But that’s not all: You can even create charts beyond your initial dataset. Data Formulator automatically identifies extra computation needs, generates fields for you, and outputs the final visualization. Find the GitHub repo in the next tweet! _____ Find me → Akshay 🚀 ✔️ For more insights and tutorials on AI and Machine Learning.

Akshay 🚀

280,449 views • 1 year ago

🚨DEEPSEARCH BY xAI: THE SEARCH ENGINE THAT THINKS FOR YOU "DeepSearch is the first generation of agents that not only help engineers, researchers, and scientists with coding, but also assist everyone in answering day-to-day questions. It's like a next-generation search engine that truly helps you understand the universe." Source: xAI Elon Musk

🚨DEEPSEARCH BY xAI: THE SEARCH ENGINE THAT THINKS FOR YOU "DeepSearch is the first generation of agents that not only help engineers, researchers, and scientists with coding, but also assist everyone in answering day-to-day questions. It's like a next-generation search engine that truly helps you understand the universe." Source: xAI Elon Musk

Mario Nawfal

82,488 views • 1 year ago

NVIDIA has published a paper on DREAMGEN – a powerful 4-step pipeline for generating synthetic data for humanoids that enables task and environment generalization. - Step 1: Fine-tune a video generation model using a small number of human teleoperation videos - Step 2: Prompt the fine-tuned model to turn a single real image into new AI-imagined videos - Step 3: Automatically label actions in the generated videos - Step 4: Train a robot AI model with the labeled synthetic dataset This enabled humanoid robots to perform 22 novel behaviors – such as pouring, opening/closing articulated objects, and manipulating a variety of tools. The original teleoperation dataset only included pick-and-place tasks. This takes task extensibility to another level without requiring human teleoperation for every single task. The pipeline will be made open-source soon. Project page:

NVIDIA has published a paper on DREAMGEN – a powerful 4-step pipeline for generating synthetic data for humanoids that enables task and environment generalization. - Step 1: Fine-tune a video generation model using a small number of human teleoperation videos - Step 2: Prompt the fine-tuned model to turn a single real image into new AI-imagined videos - Step 3: Automatically label actions in the generated videos - Step 4: Train a robot AI model with the labeled synthetic dataset This enabled humanoid robots to perform 22 novel behaviors – such as pouring, opening/closing articulated objects, and manipulating a variety of tools. The original teleoperation dataset only included pick-and-place tasks. This takes task extensibility to another level without requiring human teleoperation for every single task. The pipeline will be made open-source soon. Project page:

The Humanoid Hub

12,074 views • 1 year ago

Jensen Huang says agentic AI is not a chatbot. It is a "digital robot" -- and the next bottleneck is not the model. It is first-person data. "Agentic AI is just a digital robot. It understands, it reasons, it plans, and it acts and uses tools." Then he makes the bigger jump: "The world's video data is from a third person, not first person. And so agentic systems, robotic systems, physical AI -- the data is the hardest problem." That is the hidden constraint in robotics. LLMs got the internet. Physical AI needs demonstrations, synthetic worlds, reinforcement learning, and first-person experience. This is why Jensen keeps calling it a stack: robots are not waiting for one magical model. They need the data engine, the simulator, the reward system, the runtime, and the computer. The next wave of AI is not just software that talks. It is software that sees, plans, acts, and eventually moves through the world.

Jensen Huang says agentic AI is not a chatbot. It is a "digital robot" -- and the next bottleneck is not the model. It is first-person data. "Agentic AI is just a digital robot. It understands, it reasons, it plans, and it acts and uses tools." Then he makes the bigger jump: "The world's video data is from a third person, not first person. And so agentic systems, robotic systems, physical AI -- the data is the hardest problem." That is the hidden constraint in robotics. LLMs got the internet. Physical AI needs demonstrations, synthetic worlds, reinforcement learning, and first-person experience. This is why Jensen keeps calling it a stack: robots are not waiting for one magical model. They need the data engine, the simulator, the reward system, the runtime, and the computer. The next wave of AI is not just software that talks. It is software that sees, plans, acts, and eventually moves through the world.

Karl Mehta

14,140 views • 1 month ago

The Machine That Learns The Law Behind The Data A very very interesting US Patent US10963540B2 - Physics Informed Learning Machine describes a learning system that does not begin with data alone. It begins with a physical model, usually written as a differential equation (or PDE) dx/dt = f(x,t) A normal Machine Learning model sees scattered data and tries to fit it. A physics-informed learning machine starts with a law. Then it treats the data as evidence that updates what the model believes about the physical system. For this application, I use the patent idea on NASA C-MAPSS Turbofan engine data. The machine watches multivariate telemetry from a degrading engine and infers a hidden health state that is not measured directly. From that posterior belief, it estimates the engine’s remaining useful life. In the main 3D scene, the engine lifetime is turned into a tunnel. The spiral ribbons are real sensor channels evolving over cycle-time. The glowing core is the inferred health state. The surrounding cloud is uncertainty. The orange wall ahead is the predicted failure horizon. So the big picture is: sensor evidence comes in, posterior belief tightens, and the machine moves from uncertainty toward a concrete failure prediction. The inset posteriors make that explicit. The health posterior shows where the model believes the hidden engine condition sits at the current moment, and how sharply it believes it. The RUL posterior shows the same idea for remaining life... early on it is broad, later it shifts left and narrows as the machine becomes more certain about how close failure is. This idea is not limited to engines. The same idea can apply to data centers, CPUs, GPUs, cooling systems, power grids, robotics, batteries, and any machine that produces telemetry while obeying physical constraints. In an age where machine learning runs on massive hardware infrastructure, this kind of model matters: it can turn noisy sensor streams into early warnings before expensive systems fail.

The Machine That Learns The Law Behind The Data A very very interesting US Patent US10963540B2 - Physics Informed Learning Machine describes a learning system that does not begin with data alone. It begins with a physical model, usually written as a differential equation (or PDE) dx/dt = f(x,t) A normal Machine Learning model sees scattered data and tries to fit it. A physics-informed learning machine starts with a law. Then it treats the data as evidence that updates what the model believes about the physical system. For this application, I use the patent idea on NASA C-MAPSS Turbofan engine data. The machine watches multivariate telemetry from a degrading engine and infers a hidden health state that is not measured directly. From that posterior belief, it estimates the engine’s remaining useful life. In the main 3D scene, the engine lifetime is turned into a tunnel. The spiral ribbons are real sensor channels evolving over cycle-time. The glowing core is the inferred health state. The surrounding cloud is uncertainty. The orange wall ahead is the predicted failure horizon. So the big picture is: sensor evidence comes in, posterior belief tightens, and the machine moves from uncertainty toward a concrete failure prediction. The inset posteriors make that explicit. The health posterior shows where the model believes the hidden engine condition sits at the current moment, and how sharply it believes it. The RUL posterior shows the same idea for remaining life... early on it is broad, later it shifts left and narrows as the machine becomes more certain about how close failure is. This idea is not limited to engines. The same idea can apply to data centers, CPUs, GPUs, cooling systems, power grids, robotics, batteries, and any machine that produces telemetry while obeying physical constraints. In an age where machine learning runs on massive hardware infrastructure, this kind of model matters: it can turn noisy sensor streams into early warnings before expensive systems fail.

Mathelirium

17,843 views • 3 months ago

1000+ FPS 4D Gaussian Splatting for Dynamic Scene Rendering Contributions: • We delve into the temporal redundancy of 4D Gaussian Splatting and explain the main reason for the storage pressure and suboptimal rendering speed. • We introduce 4DGS-1K, a compact and memory-efficient framework to address these issues. It consists of two key components: a spatial-temporal variation score-based pruning strategy and a temporal filter. • Extensive experiments demonstrate that 4DGS-1K not only achieves a substantial storage reduction of approximately 41× but also accelerates rasterization to 1000+ FPS while maintaining high-quality reconstruction.

1000+ FPS 4D Gaussian Splatting for Dynamic Scene Rendering Contributions: • We delve into the temporal redundancy of 4D Gaussian Splatting and explain the main reason for the storage pressure and suboptimal rendering speed. • We introduce 4DGS-1K, a compact and memory-efficient framework to address these issues. It consists of two key components: a spatial-temporal variation score-based pruning strategy and a temporal filter. • Extensive experiments demonstrate that 4DGS-1K not only achieves a substantial storage reduction of approximately 41× but also accelerates rasterization to 1000+ FPS while maintaining high-quality reconstruction.

MrNeRF

12,366 views • 1 year ago

Egocentric Videos for Robot Training 🧵 Researchers at NYU and UC Berkeley published research where they developed a system called EgoZero that trains robots using human demonstration videos recorded with smart glasses. 👓 The system converts first-person human actions into 3D point-based state-action representations. The policies executed on a gripper-equipped robot, achieved a 70% zero-shot success rate across seven manipulation tasks, with only 20 minutes of human data per task. EgoZero stands as one of the solid empirical proofs that egocentric smart-glass video collected from everyday human behavior can serve as powerful, scalable training data for real robot learning. Vincent Liu Ademi Adeniji

Egocentric Videos for Robot Training 🧵 Researchers at NYU and UC Berkeley published research where they developed a system called EgoZero that trains robots using human demonstration videos recorded with smart glasses. 👓 The system converts first-person human actions into 3D point-based state-action representations. The policies executed on a gripper-equipped robot, achieved a 70% zero-shot success rate across seven manipulation tasks, with only 20 minutes of human data per task. EgoZero stands as one of the solid empirical proofs that egocentric smart-glass video collected from everyday human behavior can serve as powerful, scalable training data for real robot learning. Vincent Liu Ademi Adeniji

VaderResearch

23,055 views • 11 months ago