No Classifier Left Behind: An In-depth Study of the RBF SVM Classifier's Vulnerability to Image Extraction Attacks via Confidence Information Exploitation.
Michael R. ClarkPeter SwartzAndrew AltenRaed M. SalihPublished in: CogMI (2020)
Keyphrases
- svm classifier
- support vector machine
- image classification
- support vector machine svm
- confidence scores
- information extraction
- image features
- training set
- support vector
- multi class
- linear classifiers
- input image
- artificial neural networks
- kernel function
- neural network
- final classification
- decision function
- svm classification
- fold cross validation
- binary classifiers
- support vectors
- hyperplane
- basis functions
- learning algorithm
- feature selection
- training data
- feature set