Back to Search

Quaternion capsule networks

Title	Quaternion capsule networks
Author	Özcan, Barış, Kınlı, Osman Furkan
Publication Date:	2021
Publication Place	- IEEE
Type	Document
Language	English
Digital	Yes
Manuscript	No
Library:	Özyeğin University
Library Asset ID	978-1-7281-8808-9
Record ID	1d698c85-b393-4673-a82b-ee4ca992bbe1
Library Location	Computer Science
Date	2021
Sample Text	Capsules are grouping of neurons that allow to represent sophisticated information of a visual entity such as pose and features. In the view of this property, Capsule Networks outperform CNNs in challenging tasks like object recognition in unseen viewpoints, and this is achieved by learning the transformations between the object and its parts with the help of high dimensional representation of pose information. In this paper, we present Quaternion Capsules (QCN) where pose information of capsules and their transformations are represented by quaternions. Quaternions arc immune to the gimbal lock, have straightforward regularization of the rotation representation for capsules, and require less number of parameters than matrices. The experimental results show that QCNs generalize better to novel viewpoints with fewer parameters, and also achieve onpar or better performances with the state-of-the-art Capsule architectures on well-known benchmarking datasets. Our code is available(1).
DOI	10.1109/ICPR48806.2021.9412006

View in source Özyeğin University Özyeğin University - Historical works, archives, and periodicals search engine

Özyeğin University

Quaternion capsule networks

Author Özcan, Barış, Kınlı, Osman Furkan

Publication Date 2021

Publication Place - IEEE

Type Document

Language English

Digital Yes

Manuscript No

Library Özyeğin University

Library Asset ID 978-1-7281-8808-9

Record ID 1d698c85-b393-4673-a82b-ee4ca992bbe1

Library Location Computer Science

Date 2021

Sample Text Capsules are grouping of neurons that allow to represent sophisticated information of a visual entity such as pose and features. In the view of this property, Capsule Networks outperform CNNs in challenging tasks like object recognition in unseen viewpoints, and this is achieved by learning the transformations between the object and its parts with the help of high dimensional representation of pose information. In this paper, we present Quaternion Capsules (QCN) where pose information of capsules and their transformations are represented by quaternions. Quaternions arc immune to the gimbal lock, have straightforward regularization of the rotation representation for capsules, and require less number of parameters than matrices. The experimental results show that QCNs generalize better to novel viewpoints with fewer parameters, and also achieve onpar or better performances with the state-of-the-art Capsule architectures on well-known benchmarking datasets. Our code is available(1).