Binding Actions to Objects in World Models

Ondrej Biza, Robert Platt, Jan-Willem van de Meent, Lawson L.S. Wong, Thomas Kipf

Keywords: world models, attention, binding, state factorization, atari, robotic manipulation

TL;DR: We present two attention mechanisms for binding actions to objects that help world models to factor states and to learn the physics of robotic pick-and-place.

Abstract: We study the problem of binding actions to objects in object-factored world models using action-attention mechanisms. We propose two attention mechanisms for binding actions to objects, soft attention and hard attention, which we evaluate in the context of structured world models for five environments. Our experiments show that hard attention helps contrastively-trained structured world models to learn to separate individual objects in an object-based grid-world environment. Further, we show that soft attention increases performance of factored world models trained on a robotic manipulation task. The learned action attention weights can be used to interpret the factored world model as the attention focuses on the manipulated object in the environment.

Community Implementations: [ 1 code implementation](https://www.catalyzex.com/paper/arxiv:2204.13022/code)