Description
Comparing robot imitation learning methods is hard when every paper has different code and environments. RoboManipBaselines is a framework integrating imitation learning methods and benchmark environments for robot manipulation.
It unifies data collection, training and evaluation across simulation and real robots.
Methods:Unified.
Benchmarks:Fair comparison.
Collection:Teleoperation.
Sim and real:Both.
It unifies data collection, training and evaluation across simulation and real robots.
Features
Methods:Unified.
Benchmarks:Fair comparison.
Collection:Teleoperation.
Sim and real:Both.
