Natural Synthetic Anomalies for Self-Supervised Anomaly Detection and Localization

Hannah M. Schlüter, Jeremy Tan, Benjamin Hou, Bernhard Kainz
9/30/2021
cs.CV

Official code: hmsch/natural-synthetic-anomalies (74 stars, MIT, last commit 2023-04-20).

Abstract

We introduce a simple and intuitive self-supervision task, Natural Synthetic Anomalies (NSA), for training an end-to-end model for anomaly detection and localization using only normal training data. NSA integrates Poisson image editing to seamlessly blend scaled patches of various sizes from separate images. This creates a wide range of synthetic anomalies which are more similar to natural sub-image irregularities than previous data-augmentation strategies for self-supervised anomaly detection. We evaluate the proposed method using natural and medical images. Our experiments with the MVTec AD dataset show that a model trained to localize NSA anomalies generalizes well to detecting real-world a priori unknown types of manufacturing defects. Our method achieves an overall detection AUROC of 97.2 outperforming all previous methods that learn without the use of additional datasets. Code available at https://github.com/hmsch/natural-synthetic-anomalies.

View on arXivView PDF

Code Implementations(1)

hmsch/natural-synthetic-anomaliesOfficial100%
MIT

Code for ECCV 2022 paper "Natural Synthetic Anomalies for Self-Supervised Anomaly Detection and Localization".

748Python, Jupyter NotebookJun 4, 20213 years agoMIT
anomaly-detectionself-supervised-learning

Similar Papers

Achieving state-of-the-art performance in the Medical Out-of-Distribution (MOOD) challenge using plausible synthetic anomalies

91% match2023

GeneralAD: Anomaly Detection Across Domains by Attending to Distorted Features

90% match1 repo592024

Inter-Realization Channels: Unsupervised Anomaly Detection Beyond One-Class Classification

89% match2023

Progressive Boundary Guided Anomaly Synthesis for Industrial Anomaly Detection

89% match1 repo362024

Toward Generalist Anomaly Detection via In-Context Residual Learning with Few-Shot Sample Prompts

88% match2 repos2022024

Reconstruction-Based Anomaly Localization via Knowledge-Informed Self-Training

88% match2024

Cite this paper

@article{schlter2021natural,
  title  = {Natural Synthetic Anomalies for Self-Supervised Anomaly Detection and Localization},
  author = {Hannah M. Schlüter and Jeremy Tan and Benjamin Hou and Bernhard Kainz},
  year   = {2021},
  eprint = {2109.15222},
  archivePrefix = {arXiv},
  url    = {https://arxiv.org/abs/2109.15222}
}

Add a badge to your README

Shows whether official code is linked for this paper.

Code of Paper badge
[![Code of Paper](https://codeofpaper.com/badge/2109.15222.svg)](https://codeofpaper.com/paper/2109.15222)

Discussion