Novelty and Anomaly Detection

Summary

Detecting the presence of outliers, like novelties or anomalies, with respect to a particular distribution has numerous applications in computer vision and in nearly every area of data science. In (Pidhorskyi et al., 2018) we introduced a generative based approach that aims at learning the manifold of the inliers, and that efficiently computes the likelihood of a data point to have been drawn from that distribution. The manifold learning can be problematic for high dimensional data. On the one hand specialized architectures are needed, like autoencoders (Pidhorskyi et al., 2020), on the other hand, it is important to leverage the local geometry of the manifold to improve the computational efficiency (Almohsen et al., 2022).

Related Projects

Generative Probabilistic Novelty Detection with Adversarial Autoencoders

2 minute read

Stanislav Pidhorskyi, Ranya Almohsen, Donald A Adjeroh, Gianfranco Doretto

References

1. CVPRW

   Generative Probabilistic Novelty Detection With Isometric Adversarial Autoencoders Almohsen, R., Keaton, M. R., Adjeroh, D. A., and Doretto, G. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops, 2022. Oral abstract bibTeX pdf doi

   Learning the manifold of a complex distribution is a fundamental challenge for novelty or anomaly detection. We introduce a revised learning and inference procedure that takes into account a key underlying assumption made by the framework of generative probabilistic novelty detection. The traditional framework implies the ability to not only learn the manifold of the generative distribution of inliers but also to compute non-linear orthogonal projections onto this manifold from the ambient space. We augment the original training with priors that endow the model with this property, and prove that inference becomes easier and computationally more efficient. We show experimentally that the new framework leads to improved and more stable results.

   @inproceedings{almohsenKAD22cvprw,
     abbr = {CVPRW},
     author = {Almohsen, R. and Keaton, M. R. and Adjeroh, D. A. and Doretto, G.},
     booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops},
     title = {Generative Probabilistic Novelty Detection With Isometric Adversarial Autoencoders},
     year = {2022},
     month = jun,
     pages = {2002--2012},
     publisher = {IEEE},
     bib2html_pubtype = {Conferences},
     doi = {10.1109/cvprw56347.2022.00218},
     wwwnote = {Oral}
   }
   

2. CVPR

   Adversarial Latent Autoencoders Pidhorskyi, S., Adjeroh, D. A., and Doretto, G. In Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition, 2020. abstract bibTeX arXiv pdf code

   Autoencoder networks are unsupervised approaches aiming at combining generative and representational properties by learning simultaneously an encoder-generator map. Although studied extensively, the issues of whether they have the same generative power of GANs, or learn disentangled representations, have not been fully addressed. We introduce an autoencoder that tackles these issues jointly, which we call Adversarial Latent Autoencoder (ALAE). It is a general architecture that can leverage recent improvements on GAN training procedures. We designed two autoencoders: one based on a MLP encoder, and another based on a StyleGAN generator, which we call StyleALAE. We verify the disentanglement properties of both architectures. We show that StyleALAE can not only generate 1024x1024 face images with comparable quality of StyleGAN, but at the same resolution can also produce face reconstructions and manipulations based on real images. This makes ALAE the first autoencoder able to compare with, and go beyond the capabilities of a generatoronly type of architecture.

   @inproceedings{pidhorskyiAD2020cvpr,
     abbr = {CVPR},
     author = {Pidhorskyi, S. and Adjeroh, D. A. and Doretto, G.},
     booktitle = {Proceedings of the IEEE Computer Society Conference on Computer Vision
     	and Pattern Recognition},
     title = {Adversarial Latent Autoencoders},
     pages = {14104--14113},
     year = {2020},
     arxiv = {2004.04467},
     code = {https://github.com/podgorskiy/ALAE},
     bib2html_pubtype = {Conferences}
   }
   

3. NeurIPS

   Generative probabilistic novelty detection with adversarial autoencoders Pidhorskyi, S., Almohsen, R., Adjeroh, D., and Doretto, G. In Neural Information Processing Systems, 2018. abstract bibTeX arXiv pdf code

   Novelty detection is the problem of identifying whether a new data point is considered to be an inlier or an outlier. We assume that training data is available to describe only the inlier distribution. Recent approaches primarily leverage deep encoder-decoder network architectures to compute a reconstruction error that is used to either compute a novelty score or to train a one-class classifier. While we too leverage a novel network of that kind, we take a probabilistic approach and effectively compute how likely it is that a sample was generated by the inlier distribution. We achieve this with two main contributions. First, we make the computation of the novelty probability feasible because we linearize the parameterized manifold capturing the underlying structure of the inlier distribution, and show how the probability factorizes and can be computed with respect to local coordinates of the manifold tangent space. Second, we improve the training of the autoencoder network. An extensive set of results show that the approach achieves state-of-the-art performance on several benchmark datasets.

   @inproceedings{pidhorskyiAAD18neurips,
     abbr = {NeurIPS},
     author = {Pidhorskyi, S. and Almohsen, R. and Adjeroh, D. and Doretto, G.},
     title = {Generative probabilistic novelty detection with adversarial autoencoders},
     booktitle = {Neural Information Processing Systems},
     year = {2018},
     pages = {6823--6834},
     arxiv = {1807.02588},
     code = {https://github.com/podgorskiy/GPND},
     bib2html_pubtype = {Conferences}
   }