Wafer Defect Classification Using CNN Feature Clustering
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional wafer defect detection methods face performance issues due to high-dimensional data and high computational complexity in large-scale datasets, limiting their effectiveness in semiconductor fabrication.
Innovation Solution
An inspection system utilizing a classification convolutional neural network (CNN) with an encoder and decoder, combined with a clustering system, to capture and classify wafer images, encode features, and generate soft labels for training, thereby improving defect detection accuracy and efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional image classification methods are used for wafer defect detection, then defect detection can be performed, but performance deteriorates due to high-dimensional data and high computational complexity on large-scale datasets
Solution Approach 1:
The patent extracts and removes redundant information from wafer defect images through preprocessing steps including background subtraction, noise filtering, and feature selection. This extraction process isolates only the critical defect-related features, reducing the dimensional complexity of the data while preserving the essential information needed for accurate defect detection and classification.
Solution Approach 2:
The patent replaces conventional mechanical image classification approaches with a deep learning-based system using convolutional neural networks (CNNs). This substitution enables the system to automatically learn hierarchical features from raw images without manual feature engineering, significantly improving detection accuracy while the system manages computational complexity through efficient network architecture design and training strategies.
2Productivity
If conventional image classification methods with pre-determined features are used, then defect detection can be performed, but performance deteriorates due to high computational complexity on large-scale datasets
Solution Approach 1:
The patent implements preliminary action through extensive offline training of the deep learning model on large-scale datasets. During this preliminary phase, the system learns optimal feature representations and defect patterns, storing this knowledge in the trained network weights. This allows the deployed system to perform rapid real-time defect detection without repeating the computationally intensive learning process, thereby improving processing efficiency while managing computational complexity.
3Measurement precision
If clustering methods are used to group images based on similarity, then defect classification may be improved, but performance deteriorates due to high-dimensional data complexity
Solution Approach 1:
The patent extracts essential defect features from high-dimensional image data using deep learning feature extractors. These extracted features represent the most discriminative characteristics of different defect types in a compressed form. By working with these extracted features rather than raw high-dimensional pixels, the system achieves accurate defect classification while reducing data dimensionality and computational complexity.
Solution Approach 2:
The patent transforms the problem from operating on raw image pixels to operating on learned feature representations. This parameter change involves converting high-dimensional pixel data into lower-dimensional feature vectors that capture the essential variability in defect patterns. The deep learning model learns optimal parameter transformations that maintain classification accuracy while reducing dimensionality.
Data Source
AI summary
An inspection system for determining wafer defects in semiconductor fabrication may include an image capturing device to capture a wafer image and a classification convolutional neural network (CNN) to determine a classification from a plurality of classes for the captured image. Each of the plurality of classes indicates a type of a defect in the wafer. The system may also include an encoder to encode to convert a training image into a feature vector; a cluster system to cluster the feature vector to generate soft labels for the training image; and a decoder to decode the feature vector into a re-generated image. The system may also include a classification system to determine a classification from the plurality of classes for the training image. The encoder and decoder may be formed from a CNN autoencoder. The classification CNN and the CNN autoencoder may each be a deep neural network.