Representation Learning in Partially Observable Environments using Sensorimotor Prediction

TL;DR

This paper presents a method for learning compact sensory representations in partially observable environments by integrating sensorimotor information over time to improve autonomous exploration and action.

Contribution

It introduces a model that combines sensorimotor data over time to learn useful sensory representations for prediction in complex environments.

Findings

Sensorimotor prediction aids in learning effective sensory representations.

Motor and memory components are crucial for representation learning.

The approach improves prediction accuracy in noisy, ambiguous settings.

Abstract

In order to explore and act autonomously in an environment, an agent needs to learn from the sensorimotor information that is captured while acting. By extracting the regularities in this sensorimotor stream, it can learn a model of the world, which in turn can be used as a basis for action and exploration. This requires the acquisition of compact representations from a possibly high dimensional raw observation, which is noisy and ambiguous. In this paper, we learn sensory representations from sensorimotor prediction. We propose a model which integrates sensorimotor information over time, and project it in a sensory representation which is useful for prediction. We emphasize on a simple example the role of motor and memory for learning sensory representations.