GANomaly: Semi-Supervised Anomaly Detection via Adversarial Training

By Samet Akcay, Amir Atapour-Abarghouei, Toby P. Breckon. Asian Conference on Computer Vision, 2018

1,738 citations recorded by the source corpus.

Anomaly detection is a classical problem in computer vision, namely the determination of the normal from the abnormal when datasets are highly biased towards one class (normal) due to the insufficient sample size of the other class (abnormal). While this can be addressed as a supervised learning problem, a significantly more challenging problem is that of detecting the unknown/unseen anomaly case that takes us instead into the space of a one-class, semi-supervised learning paradigm. We introduce such a novel anomaly detection model, by using a conditional generative adversarial network that jointly learns the generation of high-dimensional image space and the inference of latent space. Employing encoder-decoder-encoder sub-networks in the generator network enables the model to map the input image to a lower dimension vector, which is then used to reconstruct the generated output image. The use of the additional encoder network maps this generated image to its latent representation. Minimizing the distance between these images and the latent vectors during training aids in learning the data distribution for the normal samples. As a result, a larger distance metric from this learned data distribution at inference time is indicative of an outlier from that distribution - an anomaly. Experimentation over several benchmark datasets, from varying domains, shows the model efficacy and superiority over previous state-of-the-art approaches.

Subfields: Artificial Intelligence, Biophysics, Computer Vision and Pattern Recognition

View Paper: https://www.semanticscholar.org/paper/0535625be630c6a67f4c244ebf3aa61ad088fc70

How researchers see this paper

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

In Artificial Intelligence

In Biophysics

In Computer Vision and Pattern Recognition

Citations in the corpus

BeatGAN: Anomalous Rhythm Detection using Adversarially Generated Time Series (2019, background): “We compare the performance with PCA-based anomaly detection, one-class SVM (OCSVM) for anomalous class (using the top 50 features selected by PCA method), autoencoders (AE), variational AE (VAE), AnoGAN, and Ganomaly as shown in Table 3.”
CutPaste: Self-Supervised Learning for Anomaly Detection and Localization (2021, methodology): “The initial steps have been taken with methods including autoencoding [9, 7, 25, 59], generative adversarial networks [48, 3], using pretrained models on ImageNet [38, 45, 6, 14, 43, 44], and self-supervised learning by solving different proxy tasks with augmentations [61, 47, 57, 15].”
Anomaly Detection in Video Sequence With Appearance-Motion Correspondence (2019, methodology): “The results may thus be im( Method Belleview Train GANomaly [2] 0.735 0.194 AEs + local feature [35] 0.748 0.171 AEs + global feature [35] 0.776 0.216 ALOCC D ( X ) [40] 0.734 0.182 ALOCC D ( R ( X )) [40] 0.805 0.237 Our proposed method 0.751 0.490 SSIM on appearance stream 0.830 0.798 proved by…”
A Unified Model for Multi-class Anomaly Detection (2022, methodology): “We use such a setting to simulate the real scenario, where both normal and anomalous samples contain multiple classes .”
Student-Teacher Feature Pyramid Matching for Anomaly Detection (2021, methodology): “The main weakness of these approaches comes from the excellent generalization ability of the deep models, including variational autoencoder [3], robust autoencoder [47], conditional GAN [2], and bi-directional GAN [46], which probably allows anomalous images to be faithfully reconstructed.”
Data Augmentation techniques in time series domain: a survey and taxonomy (2022, methodology): “[53] uses the GANomaly architecture[3] to process different time series data.”
MimicGAN: Robust Projection onto Image Manifolds with Corruption Mimicking (2019, methodology): “…[60, 37], etc. Additionally, an accurate parameterization of the image manifold provides a powerful regularization – sometimes referred as the “GAN prior” – for a range of problems such as defense against adversarial attacks [48, 27], anomaly detection [6, 58], and compressive sensing [52].”
Detecting semantic anomalies (2019, background): “As mentioned earlier, a small number of works in OOD detection [25, 26] have already used the hold-out-class style of tasks for evaluation.”
Multiresolution Knowledge Distillation for Anomaly Detection (2020, background): “GAN-based approaches, like AnoGan [41], f-AnoGan [40], and GANomaly [3], attempt to ﬁnd a speciﬁc latent space where the generator’s reconstructions, obtained from samplings of this space, are analogous to the normal data. f-AnoGan and GANomaly add an extra encoder to the generator to reduce inferen”
Surface Defect Detection Methods for Industrial Products: A Review (2021, methodology): “In addition, the GANomaly model (the overall structure is encoder-decoder-encoder) was proposed in the literature [94] in 2018; by comparing the difference between the latent variables obtained by the coding and the latent variables obtained by reconstructed coding, the abnormal samples are detected”
Deep Learning for Anomaly Detection (2020, methodology): “GANomaly [3] further improves the generator over the previous work by changing the generator network to an encoder-decoder-encoder network and adding two more extra loss functions.”
Skip-GANomaly: Skip Connected and Adversarially Trained Encoder-Decoder Anomaly Detection (2019, methodology): “to perform inference within the latent space is one of the novel contributions of the proposed approach compared to the previous approaches [9], [12], [13].”
DRÆM – A discriminatively trained reconstruction embedding for surface anomaly detection (2021, methodology): “Many surface anomaly detection methods focus on image reconstruction and detect anomalies based on image reconstruction error [1, 2, 5, 24, 23, 26, 31].”
ADBench: Anomaly Detection Benchmark (2022, background): “For instance, some semi-supervised models are trained only on normal samples, and detect anomalies that deviate from the normal representations learned in the training process [7, 8, 188].”
Anomaly Detection via Reverse Distillation from One-Class Embedding (2022, methodology): “Category/Method MKD [33] Ours GT [10] GN [2] US [4] PSVDD [43] DAAD [16] MF [40] PaDiM [8] CutPaste [23] Ours”