Publication details for Mr Samet AkcayAkcay, Samet, Atapour-Abarghouei, Amir & Breckon, Toby P. (2019), GANomaly: Semi-Supervised Anomaly Detection via Adversarial Training, in Jawahar, C. V., Li, Hongdong, Mori, Greg & Schindler, Konrad eds, Lecture Notes in Computer Science 11363: 14th Asian Conference on Computer Vision (ACCV). Perth, Australia, Springer, 622-637.
- Publication type: Conference Paper
- ISSN/ISBN: 0302-9743, 1611-3349, 9783030208929, 9783030208936
- DOI: 10.1007/978-3-030-20893-6_39
- Further publication details on publisher web site
- Durham Research Online (DRO) - may include full text
- View in another repository - may include full text
Author(s) from Durham
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.