Semiconductor SEM Image Defect Classification Using Supervised and Semi-Supervised Learning with Vision Transformers

TL;DR

This paper introduces a vision transformer-based approach for automatic classification of SEM wafer defect images, achieving over 90% accuracy with minimal training data, enhancing efficiency and flexibility in semiconductor defect inspection.

Contribution

It applies vision transformers and semi-supervised learning to SEM defect classification, demonstrating high accuracy with limited labeled data and potential for platform-independent inspection tools.

Findings

Achieved over 90% classification accuracy with fewer than 15 images per defect class.

Utilized transfer learning with DinoV2 for improved performance.

Demonstrated the method's potential for faster, flexible wafer defect inspection.

Abstract

Controlling defects in semiconductor processes is important for maintaining yield, improving production cost, and preventing time-dependent critical component failures. Electron beam-based imaging has been used as a tool to survey wafers in the line and inspect for defects. However, manual classification of images for these nano-scale defects is limited by time, labor constraints, and human biases. In recent years, deep learning computer vision algorithms have shown to be effective solutions for image-based inspection applications in industry. This work proposes application of vision transformer (ViT) neural networks for automatic defect classification (ADC) of scanning electron microscope (SEM) images of wafer defects. We evaluated our proposed methods on 300mm wafer semiconductor defect data from our fab in IBM Albany. We studied 11 defect types from over 7400 total images and investigated the potential of transfer learning of DinoV2 and semi-supervised learning for improved classification accuracy and efficient computation. We were able to achieve classification accuracies of over 90% with less than 15 images per defect class. Our work demonstrates the potential to apply the proposed framework for a platform agnostic in-house classification tool with faster turnaround time and flexibility.