Example of Anomaly Detection using Convolutional Variational Auto-Encoder (CVAE) [Related repository] [PyTorch Version].
Simplified VAE architecture.
'Class-1' is defined as normal and the others are defined as abnormal.
Restoration result by CVAE.
Latent vector space of training set, and reconstruction result of latent space walking.
Left figure shows latent vector space of test set. Right figure shows box plot with restoration loss of test procedure.
Latent vector space of test set, box plot with restoration loss, and histogram of restoration loss.
- Python 3.7.4
- Tensorflow 1.14.0
- Numpy 1.17.1
- Matplotlib 3.1.1
- Scikit Learn (sklearn) 0.21.3
[1] Kingma, D. P., & Welling, M. (2013). Auto-encoding variational bayes. arXiv preprint arXiv:1312.6114.
[2] Kullback Leibler divergence. Wikipedia