Towards Pixel-Wise Anomaly Location for High-Resolution PCBA \ via Self-Supervised Image Reconstruction

2025-12-19

AI Summary

Addressing the challenges of detecting minute defects (spanning only a few pixels) in high-resolution PCBA images and the scarcity of annotated data, this paper proposes HiSIR-Net, a self-supervised pixel-level anomaly localization framework. Methodologically, it introduces two key innovations: (1) a Selective Input–Reconstruction Gating mechanism (SIR-Gate) that dynamically suppresses reconstruction artifacts and false positives; and (2) a Position-Aware Region-Optimized Patch Selection (ROPS) strategy, integrating overlapping sliding-patch reconstruction with position-encoded, multi-scale feature fusion. Evaluated on our newly constructed 4K dataset SIPCBA-500 and multiple public benchmarks, HiSIR-Net achieves state-of-the-art localization accuracy, reduces false positive rate by 37%, and attains real-time inference speed (≥25 FPS), meeting industrial production line requirements.

Technology Category

• Computer Vision: Adversarial Attacks & Robustness
• Search and Optimization: Local Search
• Machine Learning: Multi-instance/Multi-view Learning

Application Category

• Search and Retrieval-Augmented AI: Web evaluation methodologies and metrics
• Graph Algorithms and Modeling for the Web: Representation, reconstruction, and subgraph or motif discovery in Web-related graphs
• Responsible Web: Machine-in-the-loop, human agency and autonomy

Abstract

Automated defect inspection of assembled Printed Circuit Board Assemblies (PCBA) is quite challenging due to the insufficient labeled data, micro-defects with just a few pixels in visually-complex and high-resolution images. To address these challenges, we present HiSIR-Net, a High resolution, Self-supervised Reconstruction framework for pixel-wise PCBA localization. Our design combines two lightweight modules that make this practical on real 4K-resolution boards: (i) a Selective Input-Reconstruction Gate (SIR-Gate) that lets the model decide where to trust reconstruction versus the original input, thereby reducing irrelevant reconstruction artifacts and false alarms; and (ii) a Region-level Optimized Patch Selection (ROPS) scheme with positional cues to select overlapping patch reconstructions coherently across arbitrary resolutions. Organically integrating these mechanisms yields clean, high-resolution anomaly maps with low false positive (FP) rate. To bridge the gap in high-resolution PCBA datasets, we further contribute a self-collected dataset named SIPCBA-500 of 500 images. We conduct extensive experiments on our SIPCBA-500 as well as public benchmarks, demonstrating the superior localization performance of our method while running at practical speed. Full code and dataset will be made available upon acceptance.

Problem

• Automated defect inspection of high-resolution PCBA images
• Pixel-wise anomaly localization with insufficient labeled data
• Reducing false alarms in micro-defect detection

Innovation

• Self-supervised reconstruction for pixel-wise anomaly localization
• Selective Input-Reconstruction Gate reduces false alarms
• Region-level Optimized Patch Selection for high-resolution images