/tf2-ganomaly

Tensorflow2 implementation of the paper GANomaly: Semi-Supervised Anomaly Detection via Adversarial Training

Primary LanguageJupyter Notebook

tf2-ganomaly

This repository contains Tensorflow 2.0 implementation of the paper GANomaly: Semi-Supervised Anomaly Detection via Adversarial Training [1], and highly reference on Pytorch implementation [2] from the author of paper.

Environment

docker image

includes tensorflow2, sklearn, tqdm, yapf

docker pull jaycase/tf2:latest

Train and Evaluate

# cifar10
python train.py --log_dir=log --anomaly=1 --dataset=cifar10 --isize=32 --nc=3
# mnist
python train.py --log_dir=log --anomaly=2 --dataset=mnist --isize=32 --nc=1

Notebooks

Reference

  • [1] Akcay S., Atapour-Abarghouei A., Breckon T.P. (2019) GANomaly: Semi-supervised Anomaly Detection via Adversarial Training. In: Jawahar C., Li H., Mori G., Schindler K. (eds) Computer Vision – ACCV 2018. ACCV 2018. Lecture Notes in Computer Science, vol 11363. Springer, Cham.
  • [2] https://github.com/samet-akcay/ganomaly