Weakly-supervised deep anomaly detection with pairwise relation learning

This paper studies a rarely explored but critical anomaly detection problem: weakly-supervised anomaly detection with limited labeled anomalies and a large unlabeled data set. This problem is very important because it (i) enables anomalyinformed modeling which helps identify anomalies of interests a...

Full description

Saved in:
Bibliographic Details
Main Authors: PANG, Guansong, HENGEL, Anton Van Den, SHEN, Chuanhua
Format: text
Language:English
Published: Institutional Knowledge at Singapore Management University 2019
Subjects:
Online Access:https://ink.library.smu.edu.sg/sis_research/7025
https://ink.library.smu.edu.sg/context/sis_research/article/8028/viewcontent/1910.13601v1.pdf
Tags: Add Tag
No Tags, Be the first to tag this record!
Institution: Singapore Management University
Language: English
Description
Summary:This paper studies a rarely explored but critical anomaly detection problem: weakly-supervised anomaly detection with limited labeled anomalies and a large unlabeled data set. This problem is very important because it (i) enables anomalyinformed modeling which helps identify anomalies of interests and address the notorious high false positives in unsupervised anomaly detection, and (ii) eliminates the reliance on large-scale and complete labeled anomaly data in fullysupervised settings. However, the problem is especially challenging since we have only limited labeled data for a single class, and moreover, the seen anomalies often cannot cover all types of anomalies (i.e., unseen anomalies). We address this problem by formulating the problem as a pairwise relation learning task. Particularly, our approach defines a two-stream ordinal regression network to learn the relation of randomly selected instance pairs, i.e., whether the instance pair contains labeled anomalies or just unlabeled data instances. The resulting model leverages both the labeled and unlabeled data to effectively augment the data and learn generalized representations of both normality and abnormality. Extensive empirical results show that our approach (i) significantly outperforms state-of-the-art competing methods in detecting both seen and unseen anomalies and (ii) is substantially more data-efficient