Prosodic Representations of Prominence Classification Neural Networks and Autoencoders Using Bottleneck Features.
Sofoklis KakourosAntti SuniJuraj SimkoMartti VainioPublished in: INTERSPEECH (2019)
Keyphrases
- neural network
- feature vectors
- feature set
- classification accuracy
- pattern recognition
- classification method
- feature extraction
- classification models
- feature space
- classification process
- denoising
- pattern classification
- benchmark datasets
- extracted features
- pattern classification problems
- decision trees
- class labels
- extracting features
- svm classifier
- feature representation
- supervised learning
- feature reduction
- feature selection algorithms
- feature subset
- rule extraction
- high dimensionality
- genetic algorithm
- image processing
- support vector
- low level
- discriminative features
- training set
- feature analysis
- handwritten digits
- artificial neural networks
- feature values
- support vector machine
- image classification
- speech recognition
- eeg signals
- classification algorithm
- image features
- support vector machine svm
- feature representations
- gabor filters
- textural features
- feature weights
- skin lesion
- machine learning algorithms
- fuzzy decision trees