cuteboyqq / attentionnet_adgan

Attention-Map-Guided-Two-stage-Anomaly-Detection-using-Hard-Augmentation-Pytorch

(Back to menu)

Anomaly Detection Network

implement from paper Attention Map-Guided Two-stage Anomaly Detection using Hard Augmentation https://arxiv.org/pdf/1805.08318.pdf , l just implement Attention Network .

The Attention Network is able to generate Attention maps that show which part of region in image is anomaly and normal,

and able to find anomaly images by loss value, it means if loss value is greater than particular loss threshold, it will regard as anomaly image

but the acc is lower than skip-Attention GANomaly, so need ADGAN to further detect anomaly images.

below image ref. from paper https://arxiv.org/pdf/1805.08318.pdf

Implement-Issue

(Back to menu)

-Generate fake image have some black images

-Train attetion failed if the input_attn = 1, but succeed when input_attn=empty

-The cutout augmentation is not real, needs to further implement paper method

-Maybe it is not easy to implement in tensorflow2

-The loss function is not same as paper, maybe result in false result

Hard-Augmentation

l just implement cutout augmgnetation...

(Back to menu)

Generator + Discriminator model

Table of contents

• Attention-Map-Guided-Two-stage-Anomaly-Detection-using-Hard-Augmentation-Pytorch
• Implement-Issue
• Hard-Augmentation
• Requirement
• implement
  • SkipCBAM-AutoEncoder
  • Self-Attention
  • AttentionNetwork
  • ADGAN
• Train-on-custom-dataset
• Train
• Test
• Lose-value-distribution
• Reference

Requirement

pip install -r requirements.txt

implement

(Back to menu)

SkipCBAM-AutoEncoder

(Back to menu)

1. implement skip-CBAM-GANomaly first, and then start adding self-Attention into skip-CBAM-GANomaly network

Self-Attention

(Back to menu)

below image is self-Attention network , ref. from paper https://arxiv.org/pdf/1805.08318.pdf

AttentionNetwork

(Back to menu)

ADGAN

(Back to menu)

Ongoing~~~~

Train-on-custom-dataset

(Back to menu)

Custom Dataset
├── test
│   ├── 0.normal
│   │   └── normal_tst_img_0.png
│   │   └── normal_tst_img_1.png
│   │   ...
│   │   └── normal_tst_img_n.png
│   ├── 1.abnormal
│   │   └── abnormal_tst_img_0.png
│   │   └── abnormal_tst_img_1.png
│   │   ...
│   │   └── abnormal_tst_img_m.png
├── train
│   ├── 0.normal
│   │   └── normal_tst_img_0.png
│   │   └── normal_tst_img_1.png
│   │   ...
│   │   └── normal_tst_img_t.png

Train

(Back to menu)

python train.py --img-dir "[train dataset dir]" --batch-size 64 --img-size 32 --epoch 20

Test

(Back to menu)

python test.py --nomal-dir "[test normal dataset dir]" --abnormal-dir "[test abnormal dataset dir]" --view-img --img-size 32

Lose-value-distribution

(Back to menu)

Blue : normal dataset

Orange : abnormal dataset

Reference

(Back to menu)

GANomaly: Semi-Supervised Anomaly Detection via Adversarial Training

https://arxiv.org/abs/1805.06725

Skip-GANomaly: Skip Connected and Adversarially Trained Encoder-Decoder Anomaly Detection

https://arxiv.org/pdf/1901.08954.pdf

Attention Map-Guided Two-stage Anomaly Detection using Hard Augmentation

https://arxiv.org/pdf/1805.08318.pdf