Adversarially Learned One-Class Classifier for Novelty Detection

By Mohammad Sabokrou, Mohammad Khalooei, Mahmood Fathy, Ehsan Adeli. 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2018

758 citations recorded by the source corpus.

Novelty detection is the process of identifying the observation(s) that differ in some respect from the training observations (the target class). In reality, the novelty class is often absent during training, poorly sampled or not well defined. Therefore, one-class classifiers can efficiently model such problems. However, due to the unavailability of data from the novelty class, training an end-to-end deep network is a cumbersome task. In this paper, inspired by the success of generative adversarial networks for training deep models in unsupervised and semi-supervised settings, we propose an end-to-end architecture for one-class classification. Our architecture is composed of two deep networks, each of which trained by competing with each other while collaborating to understand the underlying concept in the target class, and then classify the testing samples. One network works as the novelty detector, while the other supports it by enhancing the inlier samples and distorting the outliers. The intuition is that the separability of the enhanced inliers and distorted outliers is much better than deciding on the original samples. The proposed framework applies to different related applications of anomaly and outlier detection in images and videos. The results on MNIST and Caltech-256 image datasets, along with the challenging UCSD Ped2 dataset for video anomaly detection illustrate that our proposed method learns the target class effectively and is superior to the baseline and state-of-the-art methods.

Subfields: Artificial Intelligence, Molecular Biology

View Paper: https://www.semanticscholar.org/paper/8381157eae4fbf8908d0312a9642f8e69e944449

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

In Molecular Biology

Citations in the corpus

A Unifying Review of Deep and Shallow Anomaly Detection (2020, methodology): “Further variants of deep one-class classification include multimodal [145] or time-series extensions [338] and methods that employ adversarial learning [138], [141], [339] or transfer learning [139], [142].”
Improving Unsupervised Defect Segmentation by Applying Structural Similarity to Autoencoders (2018, background): “For the classification problem, a number of approaches have been proposed (Perera and Patel, 2018; Sabokrou et al., 2018).”
Multiresolution Knowledge Distillation for Anomaly Detection (2020, methodology): “Due to the unsupervised nature of the AD problem and the restricted data access, availability of just the normal data in training, the majority of methods [36, 31, 40, 18, 34] model the normal data abstraction by extracting semantically meaningful latent features.”
Generalized Out-of-Distribution Detection: A Survey (2021, positive): “Existing works leverage GANs to generate OOD training samples and force the model predictions to be uniform (Lee et al., 2018a), generate boundary samples in the low-density region (Vernekar et al., 2019), or similarly, high-conﬁdence OOD samples (Sricharan et al., 2018), or using meta-learning the ”
Generative Probabilistic Novelty Detection with Adversarial Autoencoders (2018, methodology): “Similarly to [43, 11] it has an additional adversarial component to improve generative capabilities of decoded images and a better manifold learning.”
Patch SVDD: Patch-level SVDD for Anomaly Detection and Segmentation (2020, methodology): “Auto encoder-based methods Early deep learning approaches to anomaly detection used auto encoders [10,11,12].”
Vision-based human activity recognition: a survey (2020, background): “As an example of generative models, [162] proposes an end-to-end deep learning model for abnormal activity recognition in videos.”
ADBench: Anomaly Detection Benchmark (2022, methodology): “…for AD as well; future benchmark can consider including: ( i ) OOD methods: MSP [65], energy-based EBO [104], and Mahalanobis distance-based MDS [92]; ( ii ) novelty detection methods: OCGAN [135] and Adversarial One-Class Classiﬁer [154]; and ( iii ) OSR methods: OpenGAN [79] and PROSER [204].”
PaDiM: a Patch Distribution Modeling Framework for Anomaly Detection and Localization (2020, neutral): “Neural network architectures like autoencoders (AE) [1], [9]–[11], variational autoen-coders (VAE) [3], [12]–[14] or generative adversarial networks (GAN) [15]–[17] are trained to reconstruct normal training images only.”
VT-ADL: A Vision Transformer Network for Image Anomaly Detection and Localization (2021, background): “Network architectures mostly consist of various conﬁguration of autoencoders [12], [13], [14], [15], [16] or Generative Adversarial Network (GAN) [17], [18].”
OCGAN: One-Class Novelty Detection Using GANs With Constrained Latent Representations (2019, methodology): “Various distance measures ranging from mean squared error [23] to discriminator output [22] have been used in the literature for this purpose.”
Multiple Class Novelty Detection Under Data Distribution Shift (2020, methodology): “Many problems such as countering adversarial/trojan/poison attacks [26,25,8,47], detecting novel categories [42,40,7,37,35,39] and out-of-distribution samples [13,24,10,54] etc.”
OpenGAN: Open-Set Recognition via Open Data Generation (2021, neutral): “There are multiple lines of work addressing open-set discrimination, such as anomaly detection [9, 29, 59], outlier detection [44, 39], and open-set recognition [45, 19].”