Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Unveiling the Anomalies in an Ever-Changing World: A Benchmark for Pixel-Level Anomaly Detection in Continual Learning

Nikola Bugarin, Jovana Bugaric, Manuel Barusco, Davide Dalle Pezze, Gian Antonio Susto

TL;DR

This study investigates the problem of Pixel-Level Anomaly Detection in the Continual Learning setting, where new data arrives over time and the goal is to perform well on new and old data.

Abstract

Anomaly Detection is a relevant problem in numerous real-world applications, especially when dealing with images. However, little attention has been paid to the issue of changes over time in the input data distribution, which may cause a significant decrease in performance. In this study, we investigate the problem of Pixel-Level Anomaly Detection in the Continual Learning setting, where new data arrives over time and the goal is to perform well on new and old data. We implement several state-of-the-art techniques to solve the Anomaly Detection problem in the classic setting and adapt them to work in the Continual Learning setting. To validate the approaches, we use a real-world dataset of images with pixel-based anomalies to provide a reliable benchmark and serve as a foundation for further advancements in the field. We provide a comprehensive analysis, discussing which Anomaly Detection methods and which families of approaches seem more suitable for the Continual Learning setting.

Unveiling the Anomalies in an Ever-Changing World: A Benchmark for Pixel-Level Anomaly Detection in Continual Learning

TL;DR

This study investigates the problem of Pixel-Level Anomaly Detection in the Continual Learning setting, where new data arrives over time and the goal is to perform well on new and old data.

Abstract

Anomaly Detection is a relevant problem in numerous real-world applications, especially when dealing with images. However, little attention has been paid to the issue of changes over time in the input data distribution, which may cause a significant decrease in performance. In this study, we investigate the problem of Pixel-Level Anomaly Detection in the Continual Learning setting, where new data arrives over time and the goal is to perform well on new and old data. We implement several state-of-the-art techniques to solve the Anomaly Detection problem in the classic setting and adapt them to work in the Continual Learning setting. To validate the approaches, we use a real-world dataset of images with pixel-based anomalies to provide a reliable benchmark and serve as a foundation for further advancements in the field. We provide a comprehensive analysis, discussing which Anomaly Detection methods and which families of approaches seem more suitable for the Continual Learning setting.
Paper Structure (26 sections, 3 figures, 1 table)

This paper contains 26 sections, 3 figures, 1 table.

Table of Contents

  1. Introduction
  2. Related Work
  3. Continual Learning
  4. Anomaly Detection
  5. Continual Learning and Anomaly Detection
  6. Continual Learning Approach
  7. Anomaly Detection Methods
  8. DRAEM
  9. STFPM
  10. EfficientAD
  11. Padim
  12. PatchCore
  13. CFA
  14. FastFlow
  15. Experimental Setting
  16. ...and 11 more sections

Figures (3)

  • Figure 1: Considered CL setting for the AD problem. Each task corresponds to a new item. Our AD Model must be able to detect the anomalous products (image-level) and the defects inside the image (pixel-level) of a new item while remembering to perform well on previously seen items.
  • Figure 2: Scheme of the classic Replay approach employed to solve the Anomaly Detection problem in the Continual Learning setting.
  • Figure 3: Each plot shows an AD technique tested in the CL setting. In detail, each plot shows the behavior of the f1 pixel-level metric over time, where each point represents the average performance on all the tasks seen so far. The considered strategies are the same as defined in Sec. \ref{['sec:experimental_setting']}.