This repository contains the code for generating the results of Knowledge Distillation for Anomaly Detection paper, presented at the 31th European Symposium on Artificial Neural Networks (ESANN 2023) by Adrian Alan Pol*, Ekaterina Govorkova*, Sonja Gronroos*, Nadezda Chernyavskaya, Philip Harris, Maurizio Pierini, Isobel Ojalvo and Peter Elmer. The preprint is available HERE.
Setup the environment, and install the requierments.
conda create -n "kd4ad" python=3.9.2
conda activate kd4ad
pip install -r requirements.txt
Also, install torch==2.0.0
and tensorflow==2.6.0
with the CUDA enabled (when running on GPUs).
You can run experiments as follows (example for the baseline):
mkdir results-baseline
python3 experiment-baseline.py