Weakly-Supervised Anomaly Detection for Multimodal Data Distributions




Tan, Xu; Chen, Junqi; Rahardja, Sylwan; Yang, Jiawei; Rahardja, Susanto

N/A

IEEE International Conference on Signal Processing, Communications and Computing

2024

IEEE International Conference on Signal Processing, Communication and Computing

2024 IEEE International Conference on Signal Processing, Communications and Computing (ICSPCC)

979-8-3503-6656-3

979-8-3503-6655-6

2375-8341

2837-116X

DOIhttps://doi.org/10.1109/ICSPCC62635.2024.10770302

https://ieeexplore.ieee.org/document/10770302

https://arxiv.org/pdf/2406.09147



Weakly-supervised anomaly detection can outper-form existing unsupervised methods with the assistance of a very small number of labeled anomalies, which attracts increasing attention from researchers. However, existing weakly-supervised anomaly detection methods are limited as these methods do not factor in the multimodel nature of the real-world data distribution. To mitigate this, we propose the Weakly-supervised Variational-mixture-model-based Anomaly Detector (WVAD). WVAD excels in multimodal datasets. It consists of two components: a deep variational mixture model, and an anomaly score estimator. The deep variational mixture model captures various features of the data from different clusters, then these features are delivered to the anomaly score estimator to assess the anomaly levels. Experimental results on three real-world datasets demonstrate WVAD's superiority.



Last updated on 2025-21-03 at 10:40