Deep learning model integrates CycleGAN and YOLO for PCB defect detection

This paper proposes a novel framework for Printed Circuit Board (PCB) defect detection using infrared (IR) imagery, addressing the challenge of limited IR data. The method employs CycleGAN for unpaired image-to-image translation to generate synthetic IR images from visible-light images, simulating thermal patterns. These synthetic images, combined with limited real IR data, are used to train a YOLOv8 detector, significantly improving performance in low-data scenarios. AI

Summary written by gemini-2.5-flash-lite from 1 source. How we write summaries →

IMPACT Introduces a method to overcome data scarcity in industrial inspection using generative models for synthetic data augmentation.

RANK_REASON This is a research paper detailing a novel application of deep learning models for defect detection. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.CV →

• CycleGAN
• YOLOv8
• PCB
• Haoyuan Zheng

• paper
• other