NVIDIA Cosmos3-H-Dreams: Bringing Clinical Surgery to Life Through Generative Simulation (opens in a new tab)
Explore the original at its publisher.
Practical lessons, original research, and the details that make systems work.
Explore the original at its publisher.
A measured comparison of image-based estimates for the physical properties needed by simulation assets.
Explore the original at its publisher.
Agentic AI workflows can be used to prepare and validate digital twins for physical AI systems. Agents can inspect 3D scenes, author simulation-relevant data in... Agentic AI workflows can be used to prepare and validate digital twins for physical AI systems. Agents can inspect 3D scenes, author simulation-relevant data in OpenUSD, add physics prop
Navigation enables a robot to turn perception and motion into purposeful autonomy. Unlike locomotion, which produces stable movement, navigation must be used to... Navigation enables a robot to turn perception and motion into purposeful autonomy. Unlike locomotion, which produces stable movement, navigation must be used to continuously localize the
Robots need policies that can adapt to their sensors, environments, and tasks while running on onboard computing hardware. World models offer a foundation for... Robots need policies that can adapt to their sensors, environments, and tasks while running on onboard computing hardware. World models offer a foundation for learning physical interaction
Unlike autonomous driving or industrial robotics, healthcare robotics can’t rely on internet-scale data collection or unlimited real-world experimentation.... Unlike autonomous driving or industrial robotics, healthcare robotics can’t rely on internet-scale data collection or unlimited real-world experimentation. Every demonstration requires specia
Explore the original at its publisher.
Explore the original at its publisher.
GRASP is a new gradient-based planner for learned dynamics (a “world model”) that makes long-horizon planning practical by (1) lifting the trajectory into virtual states so optimization is parallel across time, (2) adding stochasticity directly to the state iterates for exploration, and (3) reshaping gradients so actions get clean signals while we
Physical Intelligence presents RL Tokens, a compressed VLA representation used to train small actor and critic networks for precise manipulation. In four real-robot tasks, it reports faster task stages after short online practice, including Ethernet insertion with 15 minutes of robot data in a two-hour training run.
Explore the original at its publisher.
Multi-scale robot memory combines recent visual observations with textual records of longer task progress.
An encoder (optical system) maps objects to noiseless images, which noise corrupts into measurements. Our information estimator uses only these noisy measurements and a noise model to quantify how well measurements distinguish objects. Many imaging systems produce measurements that humans never see or cannot interpret directly. Your smartphone proc
In this post, I’ll introduce a reinforcement learning (RL) algorithm based on an “alternative” paradigm: divide and conquer. Unlike traditional methods, this algorithm is not based on temporal difference (TD) learning (which has scalability challenges), and scales well to long-horizon tasks. We can do Reinforcement Learning (RL) based on divide and
Explore the original at its publisher.