Automated visual inspection of CMS HGCAL silicon sensor surface using an ensemble of a deep convolutional autoencoder and classifier.
Sonja GrönroosMaurizio PieriniNadezda ChernyavskayaPublished in: Mach. Learn. Sci. Technol. (2023)
Keyphrases
- automated visual inspection
- ensemble learning
- quality control
- surface defects
- multiple classifiers
- deep learning
- classifier ensemble
- training data
- restricted boltzmann machine
- ensemble classifier
- training set
- final classification
- machine vision
- si sio
- weak learners
- individual classifiers
- silicon dioxide
- feature selection
- defect detection
- real time
- combining classifiers
- multiple classifier systems
- feature set
- space charge
- weak classifiers
- learning algorithm
- majority voting
- sensor data
- laser scanner
- ensemble methods
- random forest
- deep belief networks
- three dimensional
- pruning algorithm
- decision tree classifiers
- gallium arsenide
- support vector
- svm classifier
- class labels
- multi class
- neural network
- unsupervised feature learning
- feature space
- high speed
- base classifiers
- multi sensor
- d objects
- classification algorithm
- decision trees
- accurate classifiers
- low cost
- unsupervised learning
- ensemble members
- support vector machine
- knn
- machine learning
- high density