Iso-Dream: Isolating and Leveraging Noncontrollable Visual Dynamics in World Models

TL;DR

Iso-Dream introduces a novel reinforcement learning method that isolates controllable and noncontrollable visual dynamics in world models, enhancing decision-making in complex, real-world scenarios like autonomous driving.

Contribution

The paper proposes a new approach that improves world models by isolating different sources of visual dynamics, enabling better long-term planning and control in vision-based systems.

Findings

Iso-Dream effectively decouples controllable and noncontrollable dynamics.

It outperforms existing methods in visual control and prediction tasks.

The approach benefits long-horizon decision-making in autonomous systems.

Abstract

World models learn the consequences of actions in vision-based interactive systems. However, in practical scenarios such as autonomous driving, there commonly exists noncontrollable dynamics independent of the action signals, making it difficult to learn effective world models. To tackle this problem, we present a novel reinforcement learning approach named Iso-Dream, which improves the Dream-to-Control framework in two aspects. First, by optimizing the inverse dynamics, we encourage the world model to learn controllable and noncontrollable sources of spatiotemporal changes on isolated state transition branches. Second, we optimize the behavior of the agent on the decoupled latent imaginations of the world model. Specifically, to estimate state values, we roll-out the noncontrollable states into the future and associate them with the current controllable state. In this way, the isolation of dynamics sources can greatly benefit long-horizon decision-making of the agent, such as a self-driving car that can avoid potential risks by anticipating the movement of other vehicles. Experiments show that Iso-Dream is effective in decoupling the mixed dynamics and remarkably outperforms existing approaches in a wide range of visual control and prediction domains.