Facilitating Sim-to-Real by Intrinsic Stochasticity of Real-Time Simulation in Reinforcement Learning for Robot Manipulation.
Amir M. Soufi EnayatiRam DershanZengjie ZhangDean RichertHomayoun NajjaranPublished in: IEEE Trans. Artif. Intell. (2024)
Keyphrases
- real robot
- reinforcement learning
- real environment
- real time
- manipulation tasks
- mobile robot
- simulated robot
- vision system
- high fidelity
- robot control
- multi robot
- robotic systems
- humanoid robot
- experimental platform
- augmented reality
- robotic control
- real world
- function approximation
- motion planning
- virtual environment
- control system
- perceptual aliasing
- obstacle avoidance
- dynamic programming
- swarm robots
- robotic platform
- robot behavior
- haptic feedback
- simulation data
- human robot interaction
- changing environment
- low cost
- machine learning