Model-Based Inverse Reinforcement Learning from Visual Demonstrations
arXiv:2010.09034
Abstract
Scaling model-based inverse reinforcement learning (IRL) to real robotic manipulation tasks with unknown dynamics remains an open problem. The key challenges lie in learning good dynamics models, developing algorithms that scale to high-dimensional state-spaces and being able to learn from both visual and proprioceptive demonstrations. In this work, we present a gradient-based inverse reinforcement learning framework that utilizes a pre-trained visual dynamics model to learn cost functions when given only visual human demonstrations. The learned cost functions are then used to reproduce the demonstrated behavior via visual model predictive control. We evaluate our framework on hardware on two basic object manipulation tasks.
Accepted at the 4th Conference on Robotic Learning (CoRL 2020), Cambridge MA, USA
References in corpus (7)
- Model-Agnostic Meta-Learning for Fast Adaptation of Deep Networks
- Guided Cost Learning: Deep Inverse Optimal Control via Policy Optimization
- Visual Foresight: Model-Based Deep Reinforcement Learning for Vision-Based Robotic Control
- Dense Object Nets: Learning Dense Visual Object Descriptors By and For Robotic Manipulation
- Unsupervised Learning of Object Keypoints for Perception and Control
- Generalized Inner Loop Meta-Learning
- Unsupervised Perceptual Rewards for Imitation Learning