An Explainable Laser Welding Defect Recognition Method Based on Multi-Scale Class Activation Mapping

Tianyuan Liu, Hangbin Zheng, Jinsong Bao, Pai Zheng, Junliang Wang, Changqi Yang, Jun Gu

Research output: Journal article publicationJournal articleAcademic researchpeer-review

Abstract

Vision-based online defect recognition can provide insights for laser welding quality control systems. Although the visual signal contains richer quality information than the one-dimensional signal, the quality features contained in the visual signal are more abstract. To improve the explainability of current convolutional neural networks (CNNs) for laser welding defect recognition (LWDR), a class activation mapping method based on multi-scale fusion features (CAM-MSFF) is proposed. In addition, a multi-scale features adaptive fusion method is proposed with three steps of feature squeeze, feature mapping, and feature recalibrating. In order to facilitate the learning and utilization of multi-scale features by the proposed method, supervisory information is applied to multiple scales. The experimental results show that the proposed CAM-MSFF method has higher accuracy and convergence speed than the conventional model. The results of the explainability tests show that the proposed method can provide a more accurate and human-comprehensible explanation of the model's decision basis.

Original languageEnglish
JournalIEEE Transactions on Instrumentation and Measurement
Volume71
DOIs
Publication statusPublished - 2022

Keywords

  • Class activation mapping (CAM)
  • convolutional neural network (CNN)
  • defect recognition
  • explainable deep learning (XDL)
  • laser welding
  • multi-scale

ASJC Scopus subject areas

  • Instrumentation
  • Electrical and Electronic Engineering

Cite this