This is an early access version, the complete PDF, HTML, and XML versions will be available soon.
Open AccessArticle
Multi-Class Hypersphere Anomaly Detection Based on Edge Outlier Exposure Set and Margin
by
Min Gao
Min Gao 1,
Xuan Liu
Xuan Liu 1,
Di Xu
Di Xu 1 and
Guowei Yang
Guowei Yang 2,*
1
School of Computer Science (School of Intelligent Auditing), Nanjing Audit University, Nanjing 211815, China
2
School of Electronic Information, Qingdao University, Qingdao 266071, China
*
Author to whom correspondence should be addressed.
Mathematics 2024, 12(15), 2340; https://doi.org/10.3390/math12152340 (registering DOI)
Submission received: 19 June 2024
/
Revised: 19 July 2024
/
Accepted: 25 July 2024
/
Published: 26 July 2024
Abstract
Currently, the decision boundary of the multi-class anomaly detection algorithm based on deep learning does not sufficiently capture the positive class region, posing a risk of abnormal sample features falling into the domain of normal sample features and potentially leading to misleading outcomes in practical applications. In response to the above problems, this paper proposes a new method called multi-class hypersphere anomaly detection (MMHAD) based on the edge outlier exposure set and margin. The method aims to utilize convolutional neural networks for joint training of all normal object classes, identifying a shared set of outlier exposures, learning compact identification features, and setting appropriate edge parameters to guide the model in mapping outliers outside the hypersphere. This approach enables more comprehensive detection of various types of exceptions. The experiments demonstrate that the algorithm is superior to the most advanced baseline method, with an improvement of 26.0%, 8.2%, and 20.1% on CIFAR-10 and 14.8%, 12.0%, and 20.1% on FMNIST in the cases of (2/8), (5/5), and (9,1), respectively. Furthermore, we investigate the challenging (2/18) case on CIFAR-100, where our method achieves approximately 17.4% AUROC gain. Lastly, for a recycling waste dataset with the (4/1) case, our MMHAD yields a notable 22% enhancement in performance. Experimental results show the effectiveness of the proposed model in multi-classification anomaly detection.
Share and Cite
MDPI and ACS Style
Gao, M.; Liu, X.; Xu, D.; Yang, G.
Multi-Class Hypersphere Anomaly Detection Based on Edge Outlier Exposure Set and Margin. Mathematics 2024, 12, 2340.
https://doi.org/10.3390/math12152340
AMA Style
Gao M, Liu X, Xu D, Yang G.
Multi-Class Hypersphere Anomaly Detection Based on Edge Outlier Exposure Set and Margin. Mathematics. 2024; 12(15):2340.
https://doi.org/10.3390/math12152340
Chicago/Turabian Style
Gao, Min, Xuan Liu, Di Xu, and Guowei Yang.
2024. "Multi-Class Hypersphere Anomaly Detection Based on Edge Outlier Exposure Set and Margin" Mathematics 12, no. 15: 2340.
https://doi.org/10.3390/math12152340
Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details
here.
Article Metrics
Article metric data becomes available approximately 24 hours after publication online.