CLOUD: Contrastive Learning of Unsupervised Dynamics

10/23/2020
by   Jianren Wang, et al.
4

Developing agents that can perform complex control tasks from high dimensional observations such as pixels is challenging due to difficulties in learning dynamics efficiently. In this work, we propose to learn forward and inverse dynamics in a fully unsupervised manner via contrastive estimation. Specifically, we train a forward dynamics model and an inverse dynamics model in the feature space of states and actions with data collected from random exploration. Unlike most existing deterministic models, our energy-based model takes into account the stochastic nature of agent-environment interactions. We demonstrate the efficacy of our approach across a variety of tasks including goal-directed planning and imitation from observations. Project videos and code are at https://jianrenw.github.io/cloud/.

READ FULL TEXT

page 5

page 7

page 8

research
03/15/2021

Sample-efficient Reinforcement Learning Representation Learning with Curiosity Contrastive Forward Dynamics Model

Developing an agent in reinforcement learning (RL) that is capable of pe...
research
11/12/2018

Learning Latent Dynamics for Planning from Pixels

Planning has been very successful for control tasks with known environme...
research
06/10/2019

Self-Supervised Exploration via Disagreement

Efficient exploration is a long-standing problem in sensorimotor learnin...
research
12/04/2020

Planning from Pixels using Inverse Dynamics Models

Learning task-agnostic dynamics models in high-dimensional observation s...
research
06/23/2016

Learning to Poke by Poking: Experiential Learning of Intuitive Physics

We investigate an experiential learning paradigm for acquiring an intern...
research
03/07/2021

Robotic Visuomotor Control with Unsupervised Forward Model Learned from Videos

Learning an accurate model of the environment is essential for model-bas...
research
05/22/2018

Global Navigation Using Predictable and Slow Feature Analysis in Multiroom Environments, Path Planning and Other Control Tasks

Extended Predictable Feature Analysis (PFAx) [Richthofer and Wiskott, 20...

Please sign up or login with your details

Forgot password? Click here to reset