Synapse
⌘+K
Synapse
PulseExploreClubsResearchersJournals
Instagram
HomeClubsExplore
June 29, 2026Artificial Intelligence and Autonomous SystemsOpen Access

An ensemble deep learning approach for surface defect detection in aluminum die-cast gas meter lids

View Full Paper
Ask AI
Bookmark
Share

Authors

WQWuyang QianOAOlayinka AyorindeSCSuhao Chen

Discussion

Loading...

Member takes

Overview

Randomized trial demonstrates improved defect detection in aluminum lids, highlighting automated inspection benefits.

Key Points

  • This research aims to improve the detection of surface defects in aluminum die-cast gas meter lids using advanced deep learning techniques.
  • Implemented three deep learning architectures: CNN, ResNet-18, and Vision Transformer.
  • Conducted grid search and cross-validation for hyperparameter tuning.
  • Evaluated model performance on a large real-world dataset through ten training and testing cycles.
  • All models achieved high accuracy, precision, and recall; CNN and ResNet-18 outperformed ViT.
  • The ensemble model significantly enhanced prediction accuracy and robustness over individual models.
  • Paired t-tests confirmed the ensemble model's superior performance compared to CNN and ViT.

Cite This Study

Qian et al. (2026) studied this question.

synapsesocial.com/papers/6a420b51f91bb43ea91926fdhttps://doi.org/10.55092/aias20260006
View Full Paper
Ask AI
Bookmark
Share

Also Consider

Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context:

  1. 1Ensembling Deep Learning Models for Metal Surface Defect Classification2025
  2. 2Research progress of metal surface defect detection based on deep learning2025 · 1 citations
  3. 3Deep Learning-Driven Non-Destructive Testing: Optimizing Defect Detection with Ensemble Strategies2026
  4. 4Learning-Based Approach for Automated Surface Inspection with Industrial Tomography Imaging2024 · 7 citations
  5. 5Robust Industrial Surface Defect Detection Using Statistical Feature Extraction and Capsule Network Architectures2025 · 1 citations