Divide-and-conquer routing for learning heterogeneous individualized capsules.

Capsule Networks (CapsNets) have demonstrated an enhanced ability to capture spatial relationships and preserve hierarchical feature representations compared to Convolutional Neural Networks (CNNs). However, the dynamic routing mechanism in CapsNets introduces substantial computational costs and lim...

Full description

Saved in:
Bibliographic Details
Main Authors: Hailei Yuan, Qiang Ren
Format: Article
Language:English
Published: Public Library of Science (PLoS) 2025-01-01
Series:PLoS ONE
Online Access:https://doi.org/10.1371/journal.pone.0329202
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849340411792850944
author Hailei Yuan
Qiang Ren
author_facet Hailei Yuan
Qiang Ren
author_sort Hailei Yuan
collection DOAJ
description Capsule Networks (CapsNets) have demonstrated an enhanced ability to capture spatial relationships and preserve hierarchical feature representations compared to Convolutional Neural Networks (CNNs). However, the dynamic routing mechanism in CapsNets introduces substantial computational costs and limits scalability. In this paper, we propose a divide-and-conquer routing algorithm that groups primary capsules, enabling the model to leverage independent feature subspaces for more precise and efficient feature learning. By partitioning the primary capsules, the initialization of coupling coefficients is aligned with the hierarchical structure of the capsules, addressing the limitations of existing initialization strategies that either disrupt feature aggregation or lead to excessively small activation values. Additionally, the grouped routing mechanism simplifies the iterative process, reducing computational overhead and improving scalability. Extensive experiments on benchmark image classification datasets demonstrate that our approach consistently outperforms the original dynamic routing algorithm as well as other state-of-the-art routing strategies, resulting in improved feature learning and classification accuracy. Our code is available at: https://github.com/rqfzpy/DC-CapsNet.
format Article
id doaj-art-b9923d9b4a3e433896ca0cbe36dfc579
institution Kabale University
issn 1932-6203
language English
publishDate 2025-01-01
publisher Public Library of Science (PLoS)
record_format Article
series PLoS ONE
spelling doaj-art-b9923d9b4a3e433896ca0cbe36dfc5792025-08-20T03:43:55ZengPublic Library of Science (PLoS)PLoS ONE1932-62032025-01-01207e032920210.1371/journal.pone.0329202Divide-and-conquer routing for learning heterogeneous individualized capsules.Hailei YuanQiang RenCapsule Networks (CapsNets) have demonstrated an enhanced ability to capture spatial relationships and preserve hierarchical feature representations compared to Convolutional Neural Networks (CNNs). However, the dynamic routing mechanism in CapsNets introduces substantial computational costs and limits scalability. In this paper, we propose a divide-and-conquer routing algorithm that groups primary capsules, enabling the model to leverage independent feature subspaces for more precise and efficient feature learning. By partitioning the primary capsules, the initialization of coupling coefficients is aligned with the hierarchical structure of the capsules, addressing the limitations of existing initialization strategies that either disrupt feature aggregation or lead to excessively small activation values. Additionally, the grouped routing mechanism simplifies the iterative process, reducing computational overhead and improving scalability. Extensive experiments on benchmark image classification datasets demonstrate that our approach consistently outperforms the original dynamic routing algorithm as well as other state-of-the-art routing strategies, resulting in improved feature learning and classification accuracy. Our code is available at: https://github.com/rqfzpy/DC-CapsNet.https://doi.org/10.1371/journal.pone.0329202
spellingShingle Hailei Yuan
Qiang Ren
Divide-and-conquer routing for learning heterogeneous individualized capsules.
PLoS ONE
title Divide-and-conquer routing for learning heterogeneous individualized capsules.
title_full Divide-and-conquer routing for learning heterogeneous individualized capsules.
title_fullStr Divide-and-conquer routing for learning heterogeneous individualized capsules.
title_full_unstemmed Divide-and-conquer routing for learning heterogeneous individualized capsules.
title_short Divide-and-conquer routing for learning heterogeneous individualized capsules.
title_sort divide and conquer routing for learning heterogeneous individualized capsules
url https://doi.org/10.1371/journal.pone.0329202
work_keys_str_mv AT haileiyuan divideandconquerroutingforlearningheterogeneousindividualizedcapsules
AT qiangren divideandconquerroutingforlearningheterogeneousindividualizedcapsules