Reinforcement Learning Accelerates Humanoid Behavior Production
Reinforcement learning is used to speed the production of behavior for the Atlas humanoid robot. At the heart of the learning process is a physics-based...
Introducing the Robotics and AI Institute, formerly known as The AI Institute.
Reinforcement learning is used to speed the production of behavior for the Atlas humanoid robot. At the heart of the learning process is a physics-based...
The Robotics and AI Institute aims to accelerate robotics and artificial intelligence research by bringing together top talent and the best elements of industry and...
Introducing Theia, a vision foundation model for robotics developed by researchers at the Institute.
The annual Conference on Robot Learning (CoRL) takes place this week, November 6-9 in Munich, Germany. Read on to learn more about our accepted publications.
We present a fully autonomous real-world RL framework for mobile manipulation that can learn policies without extensive instrumentation or human supervision.
Robotic manipulation is challenging and data-driven approaches typically require large amounts of data or expert demonstrations. Therefore, we introduce a motion planner for dexterous and...
Diffusion-based policies have shown remarkable capability in executing complex robotic manipulation tasks but lack explicit characterization of geometry and semantics, which often limits their ability...
Scene representation is a crucial design choice in robotic manipulation systems. An ideal representation is expected to be 3D, dynamic, and semantic to meet the...
Recent work has shown diffusion models are an effective approach to learning the multimodal distributions arising from demonstration data in behavior cloning.
Many approaches to robot learning begin by inferring a reward function from a set of human demonstrations. To learn a good reward, it is necessary...