Year: 2020
Type: article
Abstract: Learning or identifying dynamics from a sequence of high-dimensional observations is a difficult challenge in many domains, including reinforcement learning, and control. The problem has recently been studied from a generative perspective through latent dynamics: high-dimensional observations are embedded into a lower-dimensional space in which the dynamics can be learned. Despite some successes, latent dynamics models have not yet been applied to real-world robotic systems where learned representations must be robust to a variety of perceptual confounds, and noise sources not seen during training. In this letter, we present a method to jointly learn a latent state representation, and the associated dynamics that is amenable for long-term planning, and closed-loop control under perceptually difficult conditions. As our main contribution, we describe how our representation is able to capture a notion of heteroscedastic or input-specific uncertainty at test time by detecting novel or out-of-distribution (OOD) inputs. We present results from prediction, and control experiments on two image-based tasks: a simulated pendulum balancing task, and a real-world robotic manipulator reaching task. We demonstrate that our model produces significantly more accurate predictions, and exhibits improved control performance, compared to a model that assumes homoscedastic uncertainty only, in the presence of varying degrees of input degradation.
Authors: Oliver Limoyo, Bryan Chan, Filip Maric, Brandon Wagstaff, A. Rupam Mahmood, Jonathan Kelly
Institutions: University of Toronto, University of Zagreb, University of Alberta
Language: English
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: green
Funder: Canada Research Chairs
OpenAlex ID: W3048464360