This project is an unofficial implementation of "EfficientAD: Accurate Visual Anomaly Detection at Millisecond-Level Latencies".

Datasets

./data

• ImageNet

  • n01440764
  • n01443537 ...

• MVTec_AD

  • bottle
    • ground_truth
    • test
    • train
  • cable
    • ground_truth
    • test
    • train ...

• result

Quick start

1. Install PyTorch environment

conda activate <your_env>
pip install -r requirements.txt

1. Distill a PDN architecture teacher network from wide_resnet101

python distillaion_training.py

2. train the student network and autoencoder network

python train_reduced_student.py  -c configs/mvtec_train.yaml

Pretrain Weights

1. intall Git LFS

https://git-lfs.github.com/

2. Pull LFS

cd EfficientAD
git lfs pull

Some results

Model	Dataset	Official Paper	ours
EfficientAD-M	VisA	98.1	97.54
EfficientAD-M	Mvtec LOCO	90.7	pending
EfficientAD-M	Mvtec AD	99.1	99.36
EfficientAD-S	VisA	97.5	97.20
EfficientAD-S	Mvtec LOCO	90.0	pending
EfficientAD-S	Mvtec AD	98.8	98.51

MVTec bottle