Voice-transformation-based data augmentation for prosodic classification.
Raul FernandezAndrew RosenbergAlexander SorinBhuvana RamabhadranRon HooryPublished in: ICASSP (2017)
Keyphrases
- data sets
- data processing
- pattern recognition
- database
- data sources
- image data
- input data
- data structure
- training samples
- training data
- missing values
- original data
- synthetic data
- knowledge discovery
- probability distribution
- pattern classification
- classification accuracy
- feature vectors
- high quality
- feature extraction
- feature selection
- machine learning
- active learning
- prior knowledge
- small number
- data analysis
- support vector
- data collection
- statistical analysis
- missing data
- data distribution
- decision trees
- data transformation