Grouped Discrete Representation for Object-Centric Learning

TL;DR

This paper introduces Grouped Discrete Representation (GDR), a novel approach for Object-Centric Learning that decomposes features into attribute groups and uses tuple code indexes for improved object separation and interpretability.

Contribution

GDR enhances object-centric learning by decomposing features into attribute groups and employing tuple code indexes, improving generalization and convergence.

Findings

GDR improves performance of existing OCL methods across datasets.

GDR enhances object separability and interpretability.

Visualizations show clearer object attribute distinctions.

Abstract

Object-Centric Learning (OCL) aims to discover objects in images or videos by reconstructing the input. Representative methods achieve this by reconstructing the input as its Variational Autoencoder (VAE) discrete representations, which suppress (super-)pixel noise and enhance object separability. However, these methods treat features as indivisible units, overlooking their compositional attributes, and discretize features via scalar code indexes, losing attribute-level similarities and differences. We propose Grouped Discrete Representation (GDR) for OCL. For better generalization, features are decomposed into combinatorial attributes by organized channel grouping. For better convergence, features are quantized into discrete representations via tuple code indexes. Experiments demonstrate that GDR consistently improves both mainstream and state-of-the-art OCL methods across various datasets. Visualizations further highlight GDR's superior object separability and interpretability. The source code is available on https://github.com/Genera1Z/GroupedDiscreteRepresentation.