Variational Autoencoder based Anomaly Detection using Reconstruction Probability

By Jinwon An, Sungzoon Cho. 2015

1,756 citations recorded by the source corpus.

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

Citations in the corpus

Out-of-Domain Detection for Natural Language Understanding in Dialog Systems (2019, background): “Popular approaches include modeling the probability density [8], [25], computing reconstruction losses [7], [26],”
Efficient GAN-Based Anomaly Detection (2018, methodology): “Autoencoders, variational autoencoders (An & Cho, 2015; Zhou & Paffenroth, 2017), energy based models (Zhai et al., 2016) and deep autoencoding Gaussian mixture models (Bo Zong, 2018) have been explored for anomaly detection.”
Potential, challenges and future directions for deep learning in prognostics and health management applications (2020, methodology): “Several works apply generative deep modelling, either variational autoencoders (VAE)s or GANs for anomaly detection (An and Cho, 2015; Schlegl et al., 2017; Deecke et al., 2018).”
Deep One-Class Classification (2018, methodology): “…2016; Erfani et al., 2016; Sabokrou et al., 2016), by plugging the learned embeddings into classical AD methods, but also in fully deep approaches, by directly employing the reconstruction error as an anomaly score (Hawkins et al., 2002; Sakurada & Yairi, 2014; An & Cho, 2015; Chen et al., 2017).”
Autoencoder-based feature learning for cyber security applications (2017, methodology): “The paper [24] introduced the application of a Variational AE for intrusion detection and showed that the Variational AE can perform well for both network-based intrusion detection as well as in detecting outliers.”
FastFlow: Unsupervised Anomaly Detection and Localization via 2D Normalizing Flows (2021, unknown): “1999), KDE (Bishop 2006), l2AE (Hadsell, Chopra, and LeCun 2006), VAE (An and Cho 2015), Pixel CNN (Oord et al.”
The MVTec Anomaly Detection Dataset: A Comprehensive Real-World Dataset for Unsupervised Anomaly Detection (2021, background): “The landscape of methods for unsupervised anomaly detection is diverse and many approaches have been suggested to tackle the problem (An and Cho 2015; Pereraet and Patel 219).”
Unsupervised Brain Abnormality Detection Using High Fidelity Image Reconstruction Networks (2020, background): “More recently, developments in computer vision have enabled the development of deep-learning-based models for unsupervised abnormality detection (An and Cho, 2015; Chalapathy et al., 2017).”
MVTec AD — A Comprehensive Real-World Dataset for Unsupervised Anomaly Detection (2019, background): “A popular approach [1, 7, 21] is to select an arbitrary subset of classes, re-label them as outliers, and train a novelty detection system solely on the remaining inlier classes.”
SimpleNet: A Simple Network for Image Anomaly Detection and Localization (2023, methodology): “The baselines include VAE [2], LSA [1], DSVDD [25], OCGAN [19], HRN [15], AnoGAN [27], DAAD [14], MKD [26], DisAug CLR [28], IGD [5] and RevDist [7].”
Generalized Out-of-Distribution Detection: A Survey (2021, unknown): “AE/VAE-basedmodelscombinereconstruction-errorwith AE/VAE models (Chen et al., 2018; An & Cho, 2015) and use strategies like reconstructing by memorized normality (Gong et al., 2019; Park et al., 2020), adapting model architectures (Lai et al., 2020), and partial/conditional reconstruction (Yan et…”
Unsupervised Anomaly Detection via Variational Auto-Encoder for Seasonal KPIs in Web Applications (2018, methodology): “Although the reconstruction probability Eqφ (z |x) [logpθ (x|z)] has been adopted in [2, 33], how it actually works has not yet been made clear.”
Deep Autoencoding Models for Unsupervised Anomaly Segmentation in Brain MR Images (2018, background): “Very recently, first attempts have also been made with deep generative models such as Variational Autoencoders[8,1] (VAEs), however limited to dense neural networks and 1D data.”
Widar3.0: Zero-Effort Cross-Domain Gesture Recognition With Wi-Fi (2021, methodology): “3) The Variational Autoencoder (VAE) method [63] trains a VAE model on resident samples and employs the reconstruction loss during testing as outlier indicator.”
Improving Unsupervised Defect Segmentation by Applying Structural Similarity to Autoencoders (2018, methodology): “…approach for utilizing the posterior Q ( z | x ) that yields a spatial residual map is to decode N latent samples z 1 , z 2 , . . . , z N drawn from Q ( z | x ) and to evaluate the per-pixel reconstruction probability R V AE = P ( x | z 1 , z 2 , . . . , z N ) as described by An and Cho (2015).”