• English
  • Deutsch
  • Log In
    Password Login
    Research Outputs
    Fundings & Projects
    Researchers
    Institutes
    Statistics
Repository logo
Fraunhofer-Gesellschaft
  1. Home
  2. Fraunhofer-Gesellschaft
  3. Konferenzschrift
  4. Optimal human labelling for anomaly detection in industrial inspection
 
  • Details
  • Full
Options
2022
Conference Paper
Title

Optimal human labelling for anomaly detection in industrial inspection

Abstract
Anomaly detection with machine learning in industrial inspection systems for manufactured products relies on labelled data. This rises the question how the labelling by humans should be conducted. We consider the case where we want to optimize the cost of the combined inspection process done by humans and an algorithm. This also influences the combined performance of the trained model as well as the knowledge of the performance of this model. We focus on so called one-class classification problem models which produce a continuous outlier score. We establish some cost model for human and machine combined inspection of samples. We then discuss in this cost model how to select two optimal boundaries of the outlier score where in between these two boundaries human inspection takes place. We also frame this established knowledge into an applicable algorithm.
Author(s)
Zander, Tim  
Fraunhofer-Institut für Optronik, Systemtechnik und Bildauswertung IOSB  
Pan, Ziyan
Birnstill, Pascal  
Fraunhofer-Institut für Optronik, Systemtechnik und Bildauswertung IOSB  
Beyerer, Jürgen  
Fraunhofer-Institut für Optronik, Systemtechnik und Bildauswertung IOSB  
Mainwork
Forum Bildverarbeitung 2022  
Conference
Forum Bildverarbeitung 2022  
Open Access
File(s)
Download (209.11 KB)
Rights
CC BY-NC-SA 4.0: Creative Commons Attribution-NonCommercial-ShareAlike
DOI
10.24406/publica-610
Language
English
Fraunhofer-Institut für Optronik, Systemtechnik und Bildauswertung IOSB  
Keyword(s)
  • Mathematical methods and models

  • artificial intelligence and machine learning

  • quality control

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