• English
  • Deutsch
  • Log In
    Password Login
    Research Outputs
    Fundings & Projects
    Researchers
    Institutes
    Statistics
Repository logo
Fraunhofer-Gesellschaft
  1. Home
  2. Fraunhofer-Gesellschaft
  3. Scopus
  4. Insights of Anomaly Detection: How does Polluted Training Data Influence Performance?
 
  • Details
  • Full
Options
2024
Conference Paper
Title

Insights of Anomaly Detection: How does Polluted Training Data Influence Performance?

Abstract
Anomaly detection is one of the most popular fields for computer vision in industrial applications. The idea of training machine learning only on defect-free objects saves enormous amounts of integration effort. The state of the art shows that current methods on public data sets (e.g. MVTec AD data set [1]) have already solved the problem with AUROC segmentation scores of more than 99 %. In real-world applications training data is not as "clean" as in public data sets. This work investigates the changes in detection performance when outliers end up in the training data. For this purpose, the training data is enriched step by step with images of defective objects. The AUROC score and the anomaly score is used as a quality criterion for performance measurement. We show that state of the art methods can be very robust, but that in some scenarios a draw down of 15 percentage points is possible.
Author(s)
Lehr, Jan  
Fraunhofer-Institut für Produktionsanlagen und Konstruktionstechnik IPK  
Pape, Martin
Fraunhofer-Institut für Produktionsanlagen und Konstruktionstechnik IPK  
Günther, Samuel
Fraunhofer-Institut für Produktionsanlagen und Konstruktionstechnik IPK  
Krüger, Jörg
Mainwork
Sixteenth International Conference on Machine Vision, ICMV 2023  
Conference
International Conference on Machine Vision 2023  
DOI
10.1117/12.3023184
Language
English
Fraunhofer-Institut für Produktionsanlagen und Konstruktionstechnik IPK  
Keyword(s)
  • anomaly detection

  • automated optical inspection

  • polluted training data

  • Cookie settings
  • Imprint
  • Privacy policy
  • Api
  • Contact
© 2024