Learning Robot Skills with Temporal Variational Inference

2020/06/29 by Shankar, Tanmay, Gupta, Abhinav · 5 citations
#FOS: Computer and information sciences #Machine Learning (cs.LG) #Machine Learning (stat.ML) #Robotics (cs.RO)

paper · doi:10.48550/arxiv.2006.16232

Abstract

In this paper, we address the discovery of robotic options from demonstrations in an unsupervised manner. Specifically, we present a framework to jointly learn low-level control policies and higher-level policies of how to use them from demonstrations of a robot performing various tasks. By representing options as continuous latent variables, we frame the problem of learning these options as latent variable inference. We then present a temporal formulation of variational inference based on a temporal factorization of trajectory likelihoods,that allows us to infer options in an unsupervised manner. We demonstrate the ability of our framework to learn such options across three robotic demonstration datasets.

Cited by

• Learning a Thousand Tasks in a Day
• Learning Parameterized Skills from Demonstrations
• Curriculum-guided skill learning for long-horizon robot manipulation tasks
• Implicit Safety Alignment from Crowd Preferences
• Categorical Policies: Multimodal Policy Learning and Exploration in Continuous Control

Related

• Collaborative Robot Learning from Demonstrations using Hidden Markov Model State Distribution
• Dynamics Learning with Cascaded Variational Inference for Multi-Step Manipulation
• Active inference for Robot control: A Factor Graph Approach
• Efficient transfer learning and online adaptation with latent variable models for continuous control
• Generalizing Robot Imitation Learning with Invariant Hidden Semi-Markov Models