Novelty Detection with GAN

By Mark Kliger, Shachar Fleishman. 2018

57 citations recorded by the source corpus.

The ability of a classifier to recognize unknown inputs is important for many classification-based systems. We discuss the problem of simultaneous classification and novelty detection, i.e. determining whether an input is from the known set of classes and from which specific class, or from an unknown domain and does not belong to any of the known classes. We propose a method based on the Generative Adversarial Networks (GAN) framework. We show that a multi-class discriminator trained with a generator that generates samples from a mixture of nominal and novel data distributions is the optimal novelty detector. We approximate that generator with a mixture generator trained with the Feature Matching loss and empirically show that the proposed method outperforms conventional methods for novelty detection. Our findings demonstrate a simple, yet powerful new application of the GAN framework for the task of novelty detection.

Subfields: Artificial Intelligence, Biomedical Engineering, Epidemiology

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

1. Eric Jang
2. Ian S. Fischer
3. Hyun-Jae Choi
4. Alexander A. Alemi
5. Sergey Levine
6. Joshua V. Dillon
7. Yoshua Bengio
8. Mark W. Schmidt

In Biomedical Engineering

1. Alexander A. Alemi
2. Eric Jang
3. Mark W. Schmidt
4. James J. Little

Citations in the corpus

• Out-of-Domain Detection for Natural Language Understanding in Dialog Systems (2019, neutral): “many contexts with different alias, such as “anomaly detection” [15], “one-class classification” [18], [19], “open-set recognition” [20], or “novelty detection” [21].”
• Does Your Model Know the Digit 6 Is Not a Cat? A Less Biased Evaluation of "Outlier" Detectors (2018, background): “These methods provide an abstract idea which depends on the successful training of GANs.”
• Unsupervised Class-Incremental Learning Through Confusion (2021, unknown): “2018)).”
• A Less Biased Evaluation of Out-of-distribution Sample Detectors (2018, background): “There are also ideas that rely on GANs [6, 19, 24, 32] to detect anomalies or novelty in the data.”
• Uncertainty in the Variational Information Bottleneck (2018, background): “Density estimation, whether explicit (DeVries & Taylor, 2018) or implicit (Kliger & Fleishman, 2018), has been shown to be useful for out-of-distribution detection.”
• WAIC, but Why? Generative Ensembles for Robust Anomaly Detection (2018, background): “Despite this apparent shortcoming, GAN discriminators are still applied to successfully to anomaly detection problems (Schlegl et al., 2017; Deecke et al., 2018; Kliger & Fleishman, 2018).”