Abstract:The chips contained in any electronic device are manufactured over circular silicon wafers, which are monitored by inspection machines at different production stages. Inspection machines detect and locate any defect within the wafer and return a Wafer Defect Map (WDM), i.e., a list of the coordinates where defects lie, which can be considered a huge, sparse, and binary image. In normal conditions, wafers exhibit a small number of randomly distributed defects, while defects grouped in specific patterns might indicate known or novel categories of failures in the production line. Needless to say, a primary concern of semiconductor industries is to identify these patterns and intervene as soon as possible to restore normal production conditions. Here we address WDM monitoring as an open-set recognition problem to accurately classify WDM in known categories and promptly detect novel patterns. In particular, we propose a comprehensive pipeline for wafer monitoring based on a Submanifold Sparse Convolutional Network, a deep architecture designed to process sparse data at an arbitrary resolution, which is trained on the known classes. To detect novelties, we define an outlier detector based on a Gaussian Mixture Model fitted on the latent representation of the classifier. Our experiments on a real dataset of WDMs show that directly processing full-resolution WDMs by Submanifold Sparse Convolutions yields superior classification performance on known classes than traditional Convolutional Neural Networks, which require a preliminary binning to reduce the size of the binary images representing WDMs. Moreover, our solution outperforms state-of-the-art open-set recognition solutions in detecting novelties.

Wafer Map Defect Detection and Recognition Using Joint Local and Nonlocal Linear Discriminant Analysis

Multivariate image analysis in gaussian multi-scale space for defect detection

Wafer Map Defect Patterns Semi-Supervised Classification Using Latent Vector Representation

Semi-Supervised Framework for Wafer Defect Pattern Recognition with Enhanced Labeling.

Wafer Defect Pattern Labeling and Recognition Using Semi-Supervised Learning

Machine Learning Based Wafer Defect Detection

Auto-Labeling for Pattern Recognition of Wafer Defect Maps in Semiconductor Manufacturing

Advances in machine learning and deep learning applications towards wafer map defect recognition and classification: a review

Two-Dimensional Principal Component Analysis-Based Convolutional Autoencoder for Wafer Map Defect Detection

Coarse-to-Fine Few-Shot Defect Recognition with Dynamic Weighting and Joint Metric

Diffusion maps based k-nearest-neighbor rule technique for semiconductor manufacturing process fault detection

Recognition and Classification of Mixed Defect Pattern Wafer Map Based on Multi-Path DCNN

Wafer map defect pattern detection method based on improved attention mechanism

A Deep Convolutional Neural Network for Wafer Defect Identification on an Imbalanced Dataset in Semiconductor Manufacturing Processes

AI classification of wafer map defect patterns by using dual-channel convolutional neural network

Wafer map failure pattern classification using geometric transformation-invariant convolutional neural network

Machine Learning-Based Detection Method for Wafer Test Induced Defects

Wafer Defect Localization and Classification Using Deep Learning Techniques

Convolutional Neural Network for Semiconductor Wafer Defect Detection

Study on Defect Classification in Multi-Layer Structures Based on Fisher Linear Discriminate Analysis by Using Pulsed Eddy Current Technique

Deep Open-Set Recognition for Silicon Wafer Production Monitoring