Interferobot: aligning an optical interferometer by a reinforcement learning agent
arXiv:2006.02252
Abstract
Limitations in acquiring training data restrict potential applications of deep reinforcement learning (RL) methods to the training of real-world robots. Here we train an RL agent to align a Mach-Zehnder interferometer, which is an essential part of many optical experiments, based on images of interference fringes acquired by a monocular camera. The agent is trained in a simulated environment, without any hand-coded features or a priori information about the physics, and subsequently transferred to a physical interferometer. Thanks to a set of domain randomizations simulating uncertainties in physical measurements, the agent successfully aligns this interferometer without any fine tuning, achieving a performance level of a human expert.
arXiv commentAccepted at NeurIPS 2020 (spotlight presentation)
References in corpus (14)
- Mastering Chess and Shogi by Self-Play with a General Reinforcement Learning Algorithm
- QT-Opt: Scalable Deep Reinforcement Learning for Vision-Based Robotic Manipulation
- Solving Rubik's Cube with a Robot Hand
- Learning Dexterous In-Hand Manipulation
- Agent57: Outperforming the Atari Human Benchmark
- Domain Randomization for Transferring Deep Neural Networks from Simulation to the Real World
- Breeding the optical Schroedinger's cat state
- Grasp2Vec: Learning Object Representations from Self-Supervised Grasping
- Deep Dynamics Models for Learning Dexterous Manipulation
- Alignment procedure for the VIRGO Interferometer: experimental results from the Frascati prototype
- REPLAB: A Reproducible Low-Cost Arm Benchmark Platform for Robotic Learning
- Dexterous Manipulation with Deep Reinforcement Learning: Efficient, General, and Low-Cost
- ROBEL: Robotics Benchmarks for Learning with Low-Cost Robots
- Robust Visual Domain Randomization for Reinforcement Learning
Cited by in corpus (4)
- Reinforcement learning for optimization of variational quantum circuit architectures
- Bayesian Optimization with High-Dimensional Outputs
- Playing optical tweezers with deep reinforcement learning: in virtual, physical and augmented environments
- Adaptation of Quadruped Robot Locomotion with Meta-Learning