Mixture of experts with convolutional and variational autoencoders for anomaly detection
- PDF / 2,185,110 Bytes
- 14 Pages / 595.224 x 790.955 pts Page_size
- 50 Downloads / 196 Views
Mixture of experts with convolutional and variational autoencoders for anomaly detection Qien Yu1
· Muthu Subash Kavitha2 · Takio Kurita2
Accepted: 11 September 2020 © Springer Science+Business Media, LLC, part of Springer Nature 2020
Abstract This study focused on the problem of anomaly detection (AD) by means of mixture-of-experts network. Most of the existing AD methods solely based on the reconstruction errors or latent representation using a single low-dimensional manifold are often not ideal for the image objects with complex background. However, modeling the data as a mixture of low-dimensional nonlinear manifolds is natural and promising for the classification of anomalies. In this study to realize the promise of multi-manifold latent information for AD, we propose a mixture of experts ensemble with two convolutional variational autoencoders (CVAEs) and convolution network (MEx-CVAEC) which explicitly learns manifold relationships of data that make use of multiple encoded detections. Additionally, we integrate a linear-based CAE as a gating network which optimizes the expert structures for efficient data characterization based on the manifold of the latent space. In the expert structure the data is re-encoded after each decoder to enhance the latent detection performance and the VAE is used as a core element in the encoder-decoder-encode (EDE) pipeline. To the best of our knowledge, this is the first study suggesting a mixture of CVAEs-based models for AD. The performance of the MEx-CVAE with EDE pipeline which we names as (MEx-CVAEC) compared over two basic MEx-CVAE model with ED pipeline based on logistic regression (MEx-L) and based on CAE (MEx-C) structures. In addition, the performance of the proposed model on three different datasets show the highest average AUC value than that of the state-of-the-art for image anomalies detection task. Keywords Miture-of-experts · Convolutional and variational autoencoder · Anomaly detection · Latent detection
1 Introduction In today’s complex social environment, public security issues have become increasingly prominent and it is one of the hot issues in several countries. In recent years anomaly detection (AD) is gaining more and more attention in many applicative disciplines. It is widely used in video Takio Kurita
[email protected] Qien Yu [email protected] Muthu Subash Kavitha [email protected] 1
Department of Information Engineering, Hiroshima University, Higashi-hiroshima, Hiroshima, 739-8521, Japan
2
Graduate School of Advanced Science and Engineering, Hiroshima University, Higashi-hiroshima, Hiroshima, 739-8521, Japan
surveillance [1, 2], defect detection [3, 4], fraud detection [5], and medical imaging [6]. Recently anomaly detection also focused on vector [7] and video datasets [8]. AD is considered as the identification of instances, events, or observations that are inconsistent with expected patterns or other instances in the dataset [9–11]. This study also follows the basic definition of the AD task by using anomaly fre
Data Loading...
