Skip to main content

Task-agnostic dynamics priors for deep reinforcement learning

Author(s): Du, Y; Narasimhan, Karthik

To refer to this page use:
Abstract: Copyright 2019 by the author(s). While model-based deep reinforcement learning (RL) holds great promise for sample efficiency and generalization, learning an accurate dynamics model is often challenging and requires substantial interaction with the environment. A wide variety of domains have dynamics that share common foundations like the laws of classical mechanics, which are rarely exploited by existing algorithms. In fact, humans continuously acquire and use such dynamics priors to easily adapt to operating in new environments. In this work, we propose an approach to learn task-agnostic dynamics priors from videos and incorporate them into an RL agent. Our method involves pre-training a frame predictor on task-agnostic physics videos to initialize dynamics models (and fine-tune them) for unseen target environments. Our frame prediction architecture, SpatialNet, is designed specifically to capture localized physical phenomena and interactions. Our approach allows for both faster policy learning and convergence to better policies, outperforming competitive approaches on several different environments. We also demonstrate that incorporating this prior allows for more effective transfer between environments.
Publication Date: 1-Jan-2019
Citation: Du, Y, Narasimhan, K. (2019). Task-agnostic dynamics priors for deep reinforcement learning. 36th International Conference on Machine Learning, ICML 2019, 2019-June (3063 - 3078
Pages: 3063 - 3078
Type of Material: Journal Article
Journal/Proceeding Title: 36th International Conference on Machine Learning, ICML 2019
Version: Author's manuscript

Items in OAR@Princeton are protected by copyright, with all rights reserved, unless otherwise indicated.