Learning and Evaluating Representations for Deep One-class Classification

Abstract

We present a two-stage framework for deep one-class classification. We first learn self-supervised representations from one-class data, and then build one-class classifiers on learned representations. The framework not only allows to learn better representations, but also permits building one-class classifiers that are faithful to the target task. We argue that classifiers inspired by the statistical perspective in generative or discriminative models are more effective than existing approaches, such as a normality score from a surrogate classifier. We thoroughly evaluate different self-supervised representation learning algorithms under the proposed framework for one-class classification. Moreover, we present a novel distribution-augmented contrastive learning that extends training distributions via data augmentation to obstruct the uniformity of contrastive representations. In experiments, we demonstrate state-of-the-art performance on visual domain one-class classification benchmarks, including novelty and anomaly detection. Finally, we present visual explanations, confirming that the decision-making process of deep one-class classifiers is intuitive to humans. The code is available at https://github.com/google-research/deep_representation_one_class.

Published at International Conference on Learning Representation (ICLR) 2021. The first two authors contributed equally

References in corpus (29)

• Scikit-learn: Machine Learning in Python
• A Simple Framework for Contrastive Learning of Visual Representations
• Fashion-MNIST: a Novel Image Dataset for Benchmarking Machine Learning Algorithms
• Bootstrap your own latent: A new approach to self-supervised Learning
• Axiomatic Attribution for Deep Networks
• Conditional Image Generation with PixelCNN Decoders
• Pixel Recurrent Neural Networks
• Data-Efficient Image Recognition with Contrastive Predictive Coding
• Density estimation using Real NVP
• Sanity Checks for Saliency Maps
• A Comprehensive Survey of Data Mining-based Fraud Detection Research
• Deep Anomaly Detection with Outlier Exposure
• CSI: Novelty Detection via Contrastive Learning on Distributionally Shifted Instances
• Deep Structured Energy Based Models for Anomaly Detection
• A Theoretical Analysis of Contrastive Unsupervised Representation Learning
• Anomaly Detection using One-Class Neural Networks
• Understanding Contrastive Representation Learning through Alignment and Uniformity on the Hypersphere
• On Variational Bounds of Mutual Information
• WAIC, but Why? Generative Ensembles for Robust Anomaly Detection
• Explainable Deep One-Class Classification
• Classification-Based Anomaly Detection for General Data
• Detecting Out-of-Distribution Inputs to Deep Generative Models Using Typicality
• Contrastive Training for Improved Out-of-Distribution Detection
• Predicting What You Already Know Helps: Provable Self-Supervised Learning
• Attribute Restoration Framework for Anomaly Detection
• Why Normalizing Flows Fail to Detect Out-of-Distribution Data
• Are generative deep models for novelty detection truly better?
• Contrastive learning, multi-view redundancy, and linear models
• Density of States Estimation for Out-of-Distribution Detection

Cited by in corpus (16)

• Deep Industrial Image Anomaly Detection: A Survey
• Industrial Internet of Things Intelligence Empowering Smart Manufacturing: A Literature Review
• A Survey on Unsupervised Anomaly Detection Algorithms for Industrial Images
• Self-Supervised Anomaly Detection in Computer Vision and Beyond: A Survey and Outlook
• Dual-distribution discrepancy with self-supervised refinement for anomaly detection in medical images
• Unsupervised Outlier Detection using Memory and Contrastive Learning
• REB: Reducing Biases in Representation for Industrial Anomaly Detection
• Learning image representations for anomaly detection: application to discovery of histological alterations in drug development
• Reconstruction from edge image combined with color and gradient difference for industrial surface anomaly detection
• FAIR: Frequency-aware Image Restoration for Industrial Visual Anomaly Detection
• Evaluation of Point Pattern Features for Anomaly Detection of Defect within Random Finite Set Framework
• TracInAD: Measuring Influence for Anomaly Detection
• Constrained Contrastive Distribution Learning for Unsupervised Anomaly Detection and Localisation in Medical Images
• Towards Fine-grained Visual Representations by Combining Contrastive Learning with Image Reconstruction and Attention-weighted Pooling
• Beyond Individual Input for Deep Anomaly Detection on Tabular Data
• Contrastive Knowledge Distillation for Anomaly Detection in Multi-Illumination/Focus Display Images