提出分组离散表示，让模型更好分离物体并理解其组成属性。
Grouped Discrete Representation for Object-Centric Learning
- 将特征按通道分组分解为可组合的属性，增强表征能力
- 用元组索引量化特征，保留属性间相似性差异
- 提升主流与顶尖方法的物体分离效果，适合图像解析任务
物体中心学习（OCL）旨在通过重构输入来发现图像或视频中的物体。现有代表性方法通过变分自编码器（VAE）的离散表示进行重构，抑制（超）像素噪声并增强物体可分性。然而，这些方法将特征视为不可分割的整体，忽视其构成属性，并采用标量代码索引进行离散化，丢失了属性层面的相似性与差异性。本文提出分组离散表示（GDR）用于OCL：通过有组织的通道分组将特征分解为可组合的属性以提升泛化能力；通过元组代码索引对特征进行量化以促进收敛。实验表明，GDR在多个数据集上持续提升主流及最先进OCL方法的表现。可视化进一步证明GDR具有更优的物体可分性和可解释性。源代码已公开于https://github.com/Genera1Z/GroupedDiscreteRepresentation。
原文摘要 · Abstract (English)
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.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。