Wafer Defect Image Fusion for Accurate Deep Learning Classification
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional systems for classifying defects in semiconductor wafers fail to accurately determine defects by considering different aspects/modalities of defect images, leading to inaccuracies in defect classification.
Innovation Solution
A method and system using a deep learning network that combines multiple modalities of wafer defect images, including color, Inner Crack Imaging, and black and white images, with a reference image to focus on the defect itself, and employs a Directed Acyclic Graph (DAG) architecture to classify defects, reducing the need for labeled images and training epochs.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a single deep learning model is used for defect classification, then the system complexity is low, but the measurement precision of defects deteriorates
Solution Approach 1:
The patent divides the defect classification task into multiple specialized deep learning models, each trained on a specific imaging modality (optical, SEM, AFM). This segmentation allows each model to focus on specific defect characteristics visible in its modality, improving overall classification accuracy while maintaining manageable complexity through modular architecture
Solution Approach 2:
The patent merges the outputs of multiple deep learning models that process different imaging modalities. By combining the classification results from optical images, SEM images, and AFM images, the system achieves more accurate defect classification than any single model could achieve alone, resolving the contradiction between precision and complexity
2Measurement precision
If multiple modalities of defect images are analyzed, then the measurement precision of defects is improved, but the use of energy and processing time increases
Solution Approach 1:
The patent performs preliminary processing by training separate deep learning models on different imaging modalities before actual defect classification. During operation, pre-trained models can quickly classify defects without requiring real-time analysis of all modalities, significantly reducing processing time while maintaining high accuracy through the use of multiple image types
3Measurement precision
If multiple modalities of defect images are analyzed, then the measurement precision of defects is improved, but the productivity of the inspection system deteriorates
Solution Approach 1:
The inspection system is segmented into independent processing streams for different imaging modalities, each with its own dedicated deep learning model. This allows parallel processing of multiple image types simultaneously, maintaining high inspection throughput while achieving improved accuracy through multi-modal analysis
Data Source
AI summary
The present disclosure provides method and system 100 for classifying defects in wafer using wafer defect images, based on deep learning network. Embodiments herein uses synergy between several modalities of the wafer defect images for the classification decision. Further, by adding a mixture of modalities, information may be obtained from different sources such as color image, ICI, the black and white image, to classify the defect image. In addition to mixture of modalities, a reference image may be used for each modality. The reference image of each modality image is provided to deep learning models to concentrate on the defect itself and not on the related underlying lithography of the defect image. Further, the reference image may be provided to the training process of the deep learning models that may significantly reduce the number of labelled images and the training epochs required for convergence of the deep learning model.