Unsupervised anomaly detection in multivariate time series is important in many applications including cyber intrusion detection and medical diagnostics. Both traditional and supervised techniques had limitations due ...
详细信息
Unsupervised anomaly detection in multivariate time series is important in many applications including cyber intrusion detection and medical diagnostics. Both traditional and supervised techniques had limitations due to data scale, labeling complexity, and cluster imbalance. Also, deep learning methods have drawbacks such as sensitivity to noise and difficulty in capturing spatial-temporal correlations. To address these challenges, we propose MTSAD, a new AE-based anomaly detection model for multivariate time series data that uses ConvLSTM and transposed convolution to effectively learn spatio-temporal features. Furthermore, in this paper, we explore the effect of noise injection and data amount utilization that improves the model performance and prevents overfitting. It increases the robustness to real sensor noise and improves the robustness of anomaly detection in industrial environments. On SWaT and WADI datasets, MTSAD achieves higher F1 scores than the competing models. The results of the study also show that data amount and noise injection are very important factors that can be used to improve the performance of AE-based anomaly detection. This work offers new understandings of the optimization of reconstruction-based architectures for unsupervised multivariate time series anomaly detection.
reconstruction-based methods have achieved remarkable outcomes in unsupervised image anomaly detection by training separate models for different categories. However, when it comes to a practical unified model, these m...
详细信息
reconstruction-based methods have achieved remarkable outcomes in unsupervised image anomaly detection by training separate models for different categories. However, when it comes to a practical unified model, these methods often face the "identical shortcut'' problem, where both normal and abnormal samples can be recovered well, leading to failure in anomaly detection. To address this problem, we propose a Spatially aware reconstruction network with anomaly suppression for multi-class Anomaly Detection (SRAD). Firstly, we propose a Spatially-aware Channel Convolutional (SCC) neural network, which replaces general convolution with channel convolution and incorporates a Spatial Information Fusion (SIF) block during encoding and decoding. The SIF block is proposed to allow the model to capture rich feature representations while avoiding overemphasizing details. Secondly, to prevent the model from learning identical reconstruction, we propose an Anomaly Suppression Feedback Learning (ASFL) strategy. The ASFL strategy encourages the model to accurately reconstruct normal samples while inhibiting the reconstruction of abnormal samples through feedback mechanism. Experiments on MVTec AD, VisA and BTAD datasets demonstrate the clear superiority of SRAD compared to previous state-of-the-art unified models, e.g., achieving 98.2% I-AUROC and 97.6% P-AUROC on multi-class MVTec AD dataset. Furthermore, SRAD exhibits a high frame rate of 51 FPS on 2080 Ti GPU, making it potential for practical applications.
One mainstream of image anomaly detection is based on reconstruction. Such methods still struggle with diverse anomalies, such as near-in-distribution or deformed types. To address the challenge, we propose a Discrimi...
详细信息
Learning methods are challenged when there is not enough labelled data. It gets worse when the existing learning data have different distributions in different domains. To deal with such situations, deep unsupervised ...
详细信息
Learning methods are challenged when there is not enough labelled data. It gets worse when the existing learning data have different distributions in different domains. To deal with such situations, deep unsupervised domain adaptation techniques have newly been widely used. This study surveys such domain adaptation methods that have been used for classification tasks in computer vision. The survey includes the very recent papers on this topic that have not been included in the previous surveys and introduces a taxonomy by grouping methods published on unsupervised domain adaptation into five groups of discrepancy-, adversarial-, reconstruction-, representation-, and attention-basedmethods.
暂无评论