Automated Sewer Defects Detection Using Style-Based Generative Adversarial Networks and Fine-Tuned Well-Known CNN Classifier.
Zuxiang SituShuai TengHanlin LiuJinhua LuoQianqian ZhouPublished in: IEEE Access (2021)
Keyphrases
- fine tuned
- discriminative classifiers
- object detection
- fine tuning
- automatic detection
- detection algorithm
- false positives
- support vector machine classifier
- feature space
- discriminative learning
- domain specific
- feature set
- detection method
- training samples
- linear classifiers
- false negative
- generative model
- data driven
- automated analysis
- convolutional neural network
- decision trees
- feature selection
- detection accuracy
- network structure
- learning algorithm
- cellular neural networks
- spam detection
- classification scheme
- complex networks
- classification algorithm
- detection rate
- social networks
- support vector machine
- support vector
- boosted classifiers
- machine learning
- face recognition
- training data
- defect detection
- bayesian networks
- training set
- unsupervised learning
- semi supervised learning
- classification process
- anomaly detection