Organized Grouped Discrete Representation for Object-Centric Learning

TL;DR

This paper introduces Organized GDR (OGDR), a novel method that improves object-centric learning by correctly organizing attribute channels, leading to better object representation and segmentation performance.

Contribution

OGDR enhances GDR by organizing attribute channels, reducing redundancy, and preserving information, thus improving unsupervised object segmentation and learning.

Findings

OGDR outperforms GDR in unsupervised segmentation tasks.

OGDR improves state-of-the-art diffusion-based OCL methods.

Analysis shows OGDR reduces redundancy and preserves information better.

Abstract

Object-Centric Learning (OCL) represents dense image or video pixels as sparse object features. Representative methods utilize discrete representation composed of Variational Autoencoder (VAE) template features to suppress pixel-level information redundancy and guide object-level feature aggregation. The most recent advancement, Grouped Discrete Representation (GDR), further decomposes these template features into attributes. However, its naive channel grouping as decomposition may erroneously group channels belonging to different attributes together and discretize them as sub-optimal template attributes, which losses information and harms expressivity. We propose Organized GDR (OGDR) to organize channels belonging to the same attributes together for correct decomposition from features into attributes. In unsupervised segmentation experiments, OGDR is fully superior to GDR in augmentating classical transformer-based OCL methods; it even improves state-of-the-art diffusion-based ones. Codebook PCA and representation similarity analyses show that compared with GDR, our OGDR eliminates redundancy and preserves information better for guiding object representation learning. The source code is available in the supplementary material.