We developed a task-driven hybrid model reduction method for solving dexterous manipulation with 5 minutes of online learning.
reinforcement-learning manipulation linear-complementarity-problem model-predictive-control model-based-reinforcement-learning dexterous-robotic-hand hybrid-model-predictive-controller contact-rich-manipulation in-hand-manipulation dexterous-manipulation
-
Updated
Mar 27, 2024 - Python