Explainable Deep One-Class Classification

Abstract

Deep one-class classification variants for anomaly detection learn a mapping that concentrates nominal samples in feature space causing anomalies to be mapped away. Because this transformation is highly non-linear, finding interpretations poses a significant challenge. In this paper we present an explainable deep one-class classification method, Fully Convolutional Data Description (FCDD), where the mapped samples are themselves also an explanation heatmap. FCDD yields competitive detection performance and provides reasonable explanations on common anomaly detection benchmarks with CIFAR-10 and ImageNet. On MVTec-AD, a recent manufacturing dataset offering ground-truth anomaly maps, FCDD sets a new state of the art in the unsupervised setting. Our method can incorporate ground-truth anomaly maps during training and using even a few of these (~5) improves performance significantly. Finally, using FCDD's explanations we demonstrate the vulnerability of deep one-class classification models to spurious image features such as image watermarks.

arXiv comment25 pages, published as a conference paper at ICLR 2021

References in corpus (14)

• Fashion-MNIST: a Novel Image Dataset for Benchmarking Machine Learning Algorithms
• Axiomatic Attribution for Deep Networks
• Methods for Interpreting and Understanding Deep Neural Networks
• Explaining NonLinear Classification Decisions with Deep Taylor Decomposition
• Understanding the Effective Receptive Field in Deep Convolutional Neural Networks
• Deep Anomaly Detection with Outlier Exposure
• Deep Anomaly Detection Using Geometric Transformations
• Safe Model-based Reinforcement Learning with Stability Guarantees
• Anomaly Detection using One-Class Neural Networks
• Towards Explaining Anomalies: A Deep Taylor Decomposition of One-Class Models
• Classification-Based Anomaly Detection for General Data
• Fairwashing Explanations with Off-Manifold Detergent
• Iterative energy-based projection on a normal data manifold for anomaly localization
• Rethinking Assumptions in Deep Anomaly Detection

Cited by in corpus (20)

• Artificial Intelligence based Anomaly Detection of Energy Consumption in Buildings: A Review, Current Trends and New Perspectives
• Deep Learning for Unsupervised Anomaly Localization in Industrial Images: A Survey
• A Survey on Unsupervised Anomaly Detection Algorithms for Industrial Images
• One-Class Classification: A Survey
• Learning and Evaluating Representations for Deep One-class Classification
• Self-Supervised Anomaly Detection in Computer Vision and Beyond: A Survey and Outlook
• A Unified Survey on Anomaly, Novelty, Open-Set, and Out-of-Distribution Detection: Solutions and Future Challenges
• Unsupervised Anomaly Detection with an Enhanced Teacher for Student-Teacher Feature Pyramid Matching
• Transfer Learning Gaussian Anomaly Detection by Fine-tuning Representations
• Identifying the Defective: Detecting Damaged Grains for Cereal Appearance Inspection
• AnomalyHop: An SSL-based Image Anomaly Localization Method
• Preemptively Pruning Clever-Hans Strategies in Deep Neural Networks
• Focus Your Distribution: Coarse-to-Fine Non-Contrastive Learning for Anomaly Detection and Localization
• CVAD: A generic medical anomaly detector based on Cascade VAE
• Anomaly Detection-Based UE-Centric Inter-Cell Interference Suppression
• Beyond Individual Input for Deep Anomaly Detection on Tabular Data
• OMASGAN: Out-of-Distribution Minimum Anomaly Score GAN for Sample Generation on the Boundary
• Contrastive Knowledge Distillation for Anomaly Detection in Multi-Illumination/Focus Display Images
• Where and When: Space-Time Attention for Audio-Visual Explanations
• Data augmentation and pre-trained networks for extremely low data regimes unsupervised visual inspection