Supervised Learning and Reinforcement Learning of Feedback Models for Reactive Behaviors: Tactile Feedback Testbed
arXiv:2007.00450 · doi:10.1177/02783649221143399
Abstract
Robots need to be able to adapt to unexpected changes in the environment such that they can autonomously succeed in their tasks. However, hand-designing feedback models for adaptation is tedious, if at all possible, making data-driven methods a promising alternative. In this paper we introduce a full framework for learning feedback models for reactive motion planning. Our pipeline starts by segmenting demonstrations of a complete task into motion primitives via a semi-automated segmentation algorithm. Then, given additional demonstrations of successful adaptation behaviors, we learn initial feedback models through learning from demonstrations. In the final phase, a sample-efficient reinforcement learning algorithm fine-tunes these feedback models for novel task settings through few real system interactions. We evaluate our approach on a real anthropomorphic robot in learning a tactile feedback task.
arXiv commentAccepted for publication in the International Journal of Robotics Research (IJRR). Paper length is 22 pages (including references) with 12 figures. A video overview of the reinforcement learning experiment on the real robot can be seen at https://www.youtube.com/watch?v=yu5v-ZXo4-E. arXiv admin note: text overlap with arXiv:1710.08555
References in corpus (9)
- Continuous control with deep reinforcement learning
- TensorFlow: Large-Scale Machine Learning on Heterogeneous Distributed Systems
- Playing Atari with Deep Reinforcement Learning
- Continuous-Time Gaussian Process Motion Planning via Probabilistic Inference
- Path Integral Policy Improvement with Covariance Matrix Adaptation
- Riemannian Motion Policies
- Learning Feedback Terms for Reactive Planning and Control
- Learning Latent Space Dynamics for Tactile Servoing
- Learning Sensor Feedback Models from Demonstrations via Phase-Modulated Neural Networks