Learning to compute inner consensus: A novel approach to modeling agreement between Capsules

TL;DR

This paper introduces a novel method for learning the routing procedures in Capsule Networks, enhancing their expressiveness by allowing routing relations to be learned rather than fixed.

Contribution

It proposes two new approaches to make the routing process in Capsule Networks trainable, improving their flexibility and potential performance.

Findings

Routing procedures can be learned as network parameters.

Enhanced expressiveness of Capsule Networks.

Potential improvements in generalization and spatial information preservation.

Abstract

This project considers Capsule Networks, a recently introduced machine learning model that has shown promising results regarding generalization and preservation of spatial information with few parameters. The Capsule Network's inner routing procedures thus far proposed, a priori, establish how the routing relations are modeled, which limits the expressiveness of the underlying model. In this project, we propose two distinct ways in which the routing procedure can be learned like any other network parameter.