OCAtari: Object-Centric Atari 2600 Reinforcement Learning Environments

TL;DR

OCAtari introduces an environment extension for Atari games that enables resource-efficient object-centric state extraction, facilitating object discovery, representation learning, and reinforcement learning, thus advancing research in perceptually grounded AI.

Contribution

The paper presents OCAtari, a novel extension to Atari environments that supports object-centric perception and learning, addressing limitations of pixel-based approaches.

Findings

OCAtari effectively detects objects in Atari games.

The framework is resource-efficient for object extraction.

OCAtari enables object-centric reinforcement learning.

Abstract

Cognitive science and psychology suggest that object-centric representations of complex scenes are a promising step towards enabling efficient abstract reasoning from low-level perceptual features. Yet, most deep reinforcement learning approaches only rely on pixel-based representations that do not capture the compositional properties of natural scenes. For this, we need environments and datasets that allow us to work and evaluate object-centric approaches. In our work, we extend the Atari Learning Environments, the most-used evaluation framework for deep RL approaches, by introducing OCAtari, that performs resource-efficient extractions of the object-centric states for these games. Our framework allows for object discovery, object representation learning, as well as object-centric RL. We evaluate OCAtari's detection capabilities and resource efficiency. Our source code is available at github.com/k4ntz/OC_Atari.