SynDaCaTE: A Synthetic Dataset For Evaluating Part-Whole Hierarchical Inference

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Levi, Jake, van der Wilk, Mark
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866913906511839232
author Levi, Jake
van der Wilk, Mark
author_facet Levi, Jake
van der Wilk, Mark
contents Learning to infer object representations, and in particular part-whole hierarchies, has been the focus of extensive research in computer vision, in pursuit of improving data efficiency, systematic generalisation, and robustness. Models which are \emph{designed} to infer part-whole hierarchies, often referred to as capsule networks, are typically trained end-to-end on supervised tasks such as object classification, in which case it is difficult to evaluate whether such a model \emph{actually} learns to infer part-whole hierarchies, as claimed. To address this difficulty, we present a SYNthetic DAtaset for CApsule Testing and Evaluation, abbreviated as SynDaCaTE, and establish its utility by (1) demonstrating the precise bottleneck in a prominent existing capsule model, and (2) demonstrating that permutation-equivariant self-attention is highly effective for parts-to-wholes inference, which motivates future directions for designing effective inductive biases for computer vision.
format Preprint
id arxiv_https___arxiv_org_abs_2506_17558
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle SynDaCaTE: A Synthetic Dataset For Evaluating Part-Whole Hierarchical Inference
Levi, Jake
van der Wilk, Mark
Computer Vision and Pattern Recognition
Artificial Intelligence
Machine Learning
Learning to infer object representations, and in particular part-whole hierarchies, has been the focus of extensive research in computer vision, in pursuit of improving data efficiency, systematic generalisation, and robustness. Models which are \emph{designed} to infer part-whole hierarchies, often referred to as capsule networks, are typically trained end-to-end on supervised tasks such as object classification, in which case it is difficult to evaluate whether such a model \emph{actually} learns to infer part-whole hierarchies, as claimed. To address this difficulty, we present a SYNthetic DAtaset for CApsule Testing and Evaluation, abbreviated as SynDaCaTE, and establish its utility by (1) demonstrating the precise bottleneck in a prominent existing capsule model, and (2) demonstrating that permutation-equivariant self-attention is highly effective for parts-to-wholes inference, which motivates future directions for designing effective inductive biases for computer vision.
title SynDaCaTE: A Synthetic Dataset For Evaluating Part-Whole Hierarchical Inference
topic Computer Vision and Pattern Recognition
Artificial Intelligence
Machine Learning
url https://arxiv.org/abs/2506.17558