Boosting universal speech attributes classification with deep neural network for foreign accent characterization.
Ville HautamäkiSabato Marco SiniscalchiHamid BehravanValerio Mario SalernoIvan KukanovPublished in: INTERSPEECH (2015)
Keyphrases
- neural network
- pattern recognition
- speech recognition
- learning vector quantization
- classification accuracy
- spoken language
- feature selection
- artificial neural networks
- class labels
- decision trees
- attribute selection
- classification models
- support vector machine svm
- model selection
- classification rules
- feature vectors
- majority voting
- automatic speech recognition
- attribute values
- cost sensitive
- pattern classification
- text classification
- supervised learning
- machine learning
- classification algorithm
- generalization ability
- ensemble learning
- self organizing maps
- back propagation
- decision tree algorithm