ProtoCaps: A Fast and Non-Iterative Capsule Network Routing Method

Published: 06 Dec 2023, Last Modified: 06 Dec 2023Accepted by TMLREveryoneRevisionsBibTeX
Abstract: Capsule Networks have emerged as a powerful class of deep learning architectures, known for robust performance with relatively few parameters compared to Convolutional Neural Networks (CNNs). However, their inherent efficiency is often overshadowed by their slow, iterative routing mechanisms which establish connections between Capsule layers, posing computational challenges resulting in an inability to scale. In this paper, we introduce a novel, non-iterative routing mechanism, inspired by trainable prototype clustering. This innovative approach aims to mitigate computational complexity, while retaining, if not enhancing, performance efficacy. Furthermore, we harness a shared Capsule subspace, negating the need to project each lower-level Capsule to each higher-level Capsule, thereby significantly reducing memory requisites during training. Our approach demonstrates superior results compared to the current best non-iterative Capsule Network and tests on the Imagewoof dataset, which is too computationally demanding to handle efficiently by iterative approaches. Our findings underscore the potential of our proposed methodology in enhancing the operational efficiency and performance of Capsule Networks, paving the way for their application in increasingly complex computational scenarios. Code is available at
Submission Length: Regular submission (no more than 12 pages of main content)
Previous TMLR Submission Url:
Changes Since Last Submission: We have updated our paper in accordance to the suggestions made by the reviewers. For final camera ready version. Table has been adjusted to ensure it fits within page margins. Code link has been added to abstract. Font size in figure 1 increased from 10pt to 12pt.
Assigned Action Editor: ~Yannis_Kalantidis2
License: Creative Commons Attribution 4.0 International (CC BY 4.0)
Submission Number: 1535