Deep Nearest Neighbor Anomaly Detection

Abstract

Nearest neighbors is a successful and long-standing technique for anomaly detection. Significant progress has been recently achieved by self-supervised deep methods (e.g. RotNet). Self-supervised features however typically under-perform Imagenet pre-trained features. In this work, we investigate whether the recent progress can indeed outperform nearest-neighbor methods operating on an Imagenet pretrained feature space. The simple nearest-neighbor based-approach is experimentally shown to outperform self-supervised methods in: accuracy, few shot generalization, training time and noise robustness while making fewer assumptions on image distributions.

References in corpus (7)

• Object Detection in Optical Remote Sensing Images: A Survey and A New Benchmark
• The Unreasonable Effectiveness of Deep Features as a Perceptual Metric
• Towards K-means-friendly Spaces: Simultaneous Deep Learning and Clustering
• Deep Anomaly Detection with Outlier Exposure
• Deep Anomaly Detection Using Geometric Transformations
• Learning Representations for Automatic Colorization
• Statistical Analysis of Nearest Neighbor Methods for Anomaly Detection

Cited by in corpus (16)

• Sub-Image Anomaly Detection with Deep Pyramid Correspondences
• Omni-frequency Channel-selection Representations for Unsupervised Anomaly Detection
• 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
• ALGAN: Anomaly Detection by Generating Pseudo Anomalous Data via Latent Variables
• Machine classification for probe based quantum thermometry
• Focus Your Distribution: Coarse-to-Fine Non-Contrastive Learning for Anomaly Detection and Localization
• SLA${}^2$P: Self-supervised Anomaly Detection with Adversarial Perturbation
• GRD-Net: Generative-Reconstructive-Discriminative Anomaly Detection with Region of Interest Attention Module
• Anomaly Detection via Self-organizing Map
• Deep Anomaly Detection by Residual Adaptation
• Enhancing Unsupervised Anomaly Detection with Score-Guided Network
• Learning to Be a Transformer to Pinpoint Anomalies
• Contrastive Knowledge Distillation for Anomaly Detection in Multi-Illumination/Focus Display Images
• Data augmentation and pre-trained networks for extremely low data regimes unsupervised visual inspection