A class imbalanced wafer defect classification framework based on variational autoencoder generative adversarial network

Abstract

Wafer Defect Classification (WDC) can be crucial to the wafer fabrication process. Engineers can quickly respond to improve the technological process, averting further defects through WDC. However, due to the complex fabrication steps, wafer defects are different in various types. This causes a severe data imbalance problem in WDC. To effectively solve the problem, this study introduces a class imbalanced wafer defect classification framework (CI-WDC) based on Variational Autoencoder Generative Adversarial Network (VAE-GAN). This framework consists of VAE-GAN and wafer defect classifier. Among them, VAE-GAN is responsible for creating new samples to solve the imbalance problem while the classifier is responsible for classifying wafer defect patterns. Specifically, VAE-GAN combines the advantage of a variational autoencoder (VAE) and generative adversarial network. VAE networks can produce subtle differences that do not affect the properties of the data when generating new images. At the same time, the proposed discriminator can help us constrain the generated images to be close to real samples and avoid irrational, feature-missing, and ambiguous samples. WM-811K dataset is utilized to verify the above method. The experimental results validate that the samples generated by VAE-GAN have a significant improvement in the performance of the WDC system.

An online color and shape integrated detection method for flexible packaging surface defects

“…Chen et al [19] proposed a method to classify surface defects of printed pharmaceutical packages based on integrated classifiers, but this method required the training of a large number of data sets. Wang et al [20] and Farady et al [21] improved the classification accuracy of defects by improving the classification network, but this method had limitations on classifying tiny defects of the studied flexible packages. Deep convolutional generative adversarial network (DCGAN) is a widely used image generation network for image enhancement.…”

AI-enabled industrial equipment monitoring, diagnosis and health management

“…Wang et al [22] presented a class-imbalanced wafer defect classification framework utilizing a variational autoencoder generative adversarial network to tackle the data imbalance issue in wafer defect classification. In Lei et al [23], a fault diagnosis method for rolling bearings was proposed based on Markov transition field and multi-dimension CNN to enhance diagnostic accuracy in complex and variable working conditions with limited sample data.…”