DLR-Logo -> http://www.dlr.de
DLR Portal Home | Imprint | Privacy Policy | Contact | Deutsch
Fontsize: [-] Text [+]

Capsule Networks for Hierarchical Novelty Detection in Object Classification

de Graaff, Thies and Ribeiro de Menezes, Arthur (2022) Capsule Networks for Hierarchical Novelty Detection in Object Classification. In: 2022 IEEE Intelligent Vehicles Symposium, IV 2022, pp. 1795-1800. Institute of Electrical and Electronics Engineers. 2022 IEEE Intelligent Vehicles Symposium (IV), 2022-06-04 - 2022-06-09, Aachen, Deutschland. doi: 10.1109/IV51971.2022.9827249. ISBN 978-166548821-1.

[img] PDF - Only accessible within DLR

Official URL: https://ieeexplore.ieee.org/document/9827249


Hierarchical Novelty Detection (HND) refers to assigning labels to objects in a hierarchical category space, where non-leaf labeling represents a novelty detection of that category. By labeling a novel instance in at least one abstract category, more informed decisions can be made by an automated driving (AD) function, resulting in a safer behavior in novel situations. Current approaches are mainly composed of different architectures based on Convolutional Neural Networks (CNNs). Capsule Networks (CNs) were introduced as an alternative to CNNs that expand their capacity in tasks that were previously challenging. We explore the hierarchical nature of CNs and propose a novel approach for hierarchical novelty detection using a unified CN architecture. As a proof-of-concept, we evaluate it on a novelty detection task based on the Fashion-MNIST dataset. We define a misclassification matrix for evaluation of the performance based on a semantically sensible scenario for this dataset. The results show that our method outperforms the main CNN-based methods in the current literature in this task while also giving more flexibility for task-specific tuning and has the potential to reach state-of-the-art status in more complex HND use cases within the AD domain.

Item URL in elib:https://elib.dlr.de/188106/
Document Type:Conference or Workshop Item (Speech)
Title:Capsule Networks for Hierarchical Novelty Detection in Object Classification
AuthorsInstitution or Email of AuthorsAuthor's ORCID iDORCID Put Code
de Graaff, ThiesUNSPECIFIEDhttps://orcid.org/0009-0006-5918-9524UNSPECIFIED
Journal or Publication Title:2022 IEEE Intelligent Vehicles Symposium, IV 2022
Refereed publication:Yes
Open Access:No
Gold Open Access:No
In ISI Web of Science:Yes
Page Range:pp. 1795-1800
Publisher:Institute of Electrical and Electronics Engineers
Series Name:2022 IEEE Intelligent Vehicles Symposium (IV)
Keywords:Hierarchical Novelty Detection, Object Classification, Capsule Networks
Event Title:2022 IEEE Intelligent Vehicles Symposium (IV)
Event Location:Aachen, Deutschland
Event Type:international Conference
Event Start Date:4 June 2022
Event End Date:9 June 2022
HGF - Research field:Aeronautics, Space and Transport
HGF - Program:Transport
HGF - Program Themes:other
DLR - Research area:Transport
DLR - Program:V - no assignment
DLR - Research theme (Project):V - no assignment
Location: Oldenburg
Institutes and Institutions:Institute of Systems Engineering for Future Mobility > Systems Theory and Design
Deposited By: de Graaff, Thies
Deposited On:05 Sep 2022 15:10
Last Modified:24 Apr 2024 20:49

Repository Staff Only: item control page

Help & Contact
electronic library is running on EPrints 3.3.12
Website and database design: Copyright © German Aerospace Center (DLR). All rights reserved.