Neural Expectation Maximization

TL;DR

This paper introduces a neural network-based clustering method inspired by Expectation Maximization to discover and represent entities in perceptual grouping tasks, improving object recovery and prediction.

Contribution

It formalizes a differentiable clustering approach that learns to group and represent entities simultaneously using neural networks within a spatial mixture model.

Findings

Accurately recovers constituent objects in perceptual grouping tasks

Learned representations improve next-step prediction

Demonstrates effectiveness of neural EM framework for entity discovery

Abstract

Many real world tasks such as reasoning and physical interaction require identification and manipulation of conceptual entities. A first step towards solving these tasks is the automated discovery of distributed symbol-like representations. In this paper, we explicitly formalize this problem as inference in a spatial mixture model where each component is parametrized by a neural network. Based on the Expectation Maximization framework we then derive a differentiable clustering method that simultaneously learns how to group and represent individual entities. We evaluate our method on the (sequential) perceptual grouping task and find that it is able to accurately recover the constituent objects. We demonstrate that the learned representations are useful for next-step prediction.