Grouped Discrete Representation Guides Object-Centric Learning

TL;DR

This paper introduces Grouped Discrete Representation (GDR), a novel approach that groups features into attributes and indexes them with tuples, enhancing object-centric learning's convergence and generalization capabilities.

Contribution

The paper proposes GDR, which improves object-centric learning by grouping features into attributes and indexing them with tuples, addressing limitations of minimal units and attribute-level commonalities.

Findings

GDR consistently improves convergence across various models and datasets.

GDR enhances generalization in object-centric learning tasks.

Visualizations confirm GDR captures attribute-level information effectively.

Abstract

Similar to humans perceiving visual scenes as objects, Object-Centric Learning (OCL) can abstract dense images or videos into sparse object-level features. Transformer-based OCL handles complex textures well due to the decoding guidance of discrete representation, obtained by discretizing noisy features in image or video feature maps using template features from a codebook. However, treating features as minimal units overlooks their composing attributes, thus impeding model generalization; indexing features with natural numbers loses attribute-level commonalities and characteristics, thus diminishing heuristics for model convergence. We propose \textit{Grouped Discrete Representation} (GDR) to address these issues by grouping features into attributes and indexing them with tuple numbers. In extensive experiments across different query initializations, dataset modalities, and model architectures, GDR consistently improves convergence and generalizability. Visualizations show that our method effectively captures attribute-level information in features. The source code will be available upon acceptance.