Learning deep dynamical models from image pixels

TL;DR

This paper introduces a method combining deep auto-encoders and predictive models to learn dynamical systems directly from high-dimensional pixel observations, enabling effective system identification in complex, non-linear scenarios.

Contribution

The paper presents a novel approach that jointly learns low-dimensional embeddings and transition models from pixel data, addressing non-linear system identification challenges.

Findings

Successfully models dynamical systems from raw pixel data

Outperforms traditional linear system identification methods

Enables predictive control directly from images

Abstract

Modeling dynamical systems is important in many disciplines, e.g., control, robotics, or neurotechnology. Commonly the state of these systems is not directly observed, but only available through noisy and potentially high-dimensional observations. In these cases, system identification, i.e., finding the measurement mapping and the transition mapping (system dynamics) in latent space can be challenging. For linear system dynamics and measurement mappings efficient solutions for system identification are available. However, in practical applications, the linearity assumptions does not hold, requiring non-linear system identification techniques. If additionally the observations are high-dimensional (e.g., images), non-linear system identification is inherently hard. To address the problem of non-linear system identification from high-dimensional observations, we combine recent advances in deep learning and system identification. In particular, we jointly learn a low-dimensional embedding of the observation by means of deep auto-encoders and a predictive transition model in this low-dimensional space. We demonstrate that our model enables learning good predictive models of dynamical systems from pixel information only.