Science ❯ Artificial Intelligence ❯ Machine Learning

Reinforcement Learning

Simulation Physical AI Inference Algorithms Multi-Agent Systems World Models in AI Skill Learning Methods Teleoperation Robotics Training Online Learning AI in Robotics Simulation-Based Learning

4 ARTICLES

4mo ago

Humanoid Robot Learns Tennis From Imperfect Human Motion, Posts High Real-World Success

The open-source LATENT method builds a correctable action space from brief motion-capture fragments to transfer robust policies from simulation to the court.