Continual Learning with Out-of-Distribution Data Detection for Defect Classification

Cheng Hsueh Lin, Chia Yu Lin*, Li Jen Wang, Ted T. Kuo

*Corresponding author for this work

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

1 Scopus citations

Abstract

We propose a framework for defect detection in production lines that leverages deep learning models, out-of-distribution (OOD) detection, and continual learning to address the challenges of unknown defects and catastrophic forgetting. The proposed method divides classifier training into chronicle tasks, each introducing new defect classes and leveraging OOD detection to classify unknown defects. We evaluate the framework on a highly unbalanced product defect dataset and demonstrated that it outperformed existing approaches, improving the average F-score by 10%. Our method also improve the performance of the PODNet and DER models, but not the WA model due to its poor performance on our dataset. These results suggest that the proposed method has the potential to improve defect detection in production lines, especially for small-quantity-wide-variety production scenarios.

Original languageEnglish
Title of host publication2023 International Conference on Consumer Electronics - Taiwan, ICCE-Taiwan 2023 - Proceedings
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages337-338
Number of pages2
ISBN (Electronic)9798350324174
DOIs
StatePublished - 2023
Event2023 International Conference on Consumer Electronics - Taiwan, ICCE-Taiwan 2023 - Pingtung, Taiwan
Duration: 17 Jul 202319 Jul 2023

Publication series

Name2023 International Conference on Consumer Electronics - Taiwan, ICCE-Taiwan 2023 - Proceedings

Conference

Conference2023 International Conference on Consumer Electronics - Taiwan, ICCE-Taiwan 2023
Country/TerritoryTaiwan
CityPingtung
Period17/07/2319/07/23

Keywords

  • Continual learning
  • out-of-distribution

Fingerprint

Dive into the research topics of 'Continual Learning with Out-of-Distribution Data Detection for Defect Classification'. Together they form a unique fingerprint.

Cite this