Application Research of a Deep Learning Model Integrating CycleGAN and YOLO in PCB Infrared Defect Detection
Chao Yang, Haoyuan Zheng, Yue Ma
TL;DR
The paper tackles IR data scarcity in PCB defect detection by integrating CycleGAN-based unpaired visible-to-IR translation with YOLOv8 for defect localization. Pseudo-IR images generated from visible PCB data are combined with limited real IR samples to train a lightweight detector, achieving strong performance under low-data conditions and approaching fully supervised results. The study demonstrates that pseudo-IR augmentation preserves defect structures and thermal patterns, providing a practical, data-efficient path for industrial PCB inspection. The approach offers scalability to related cross-modal inspection tasks and highlights the value of cross-domain data synthesis in data-constrained settings.
Abstract
This paper addresses the critical bottleneck of infrared (IR) data scarcity in Printed Circuit Board (PCB) defect detection by proposing a cross-modal data augmentation framework integrating CycleGAN and YOLOv8. Unlike conventional methods relying on paired supervision, we leverage CycleGAN to perform unpaired image-to-image translation, mapping abundant visible-light PCB images into the infrared domain. This generative process synthesizes high-fidelity pseudo-IR samples that preserve the structural semantics of defects while accurately simulating thermal distribution patterns. Subsequently, we construct a heterogeneous training strategy that fuses generated pseudo-IR data with limited real IR samples to train a lightweight YOLOv8 detector. Experimental results demonstrate that this method effectively enhances feature learning under low-data conditions. The augmented detector significantly outperforms models trained on limited real data alone and approaches the performance benchmarks of fully supervised training, proving the efficacy of pseudo-IR synthesis as a robust augmentation strategy for industrial inspection.