Boosting of Prosodic and Pronunciation Features to Detect Mispronunciations of Non-Native Children.
Christian HackerTobias CincarekAndreas K. MaierAndre HeßlerElmar NöthPublished in: ICASSP (4) (2007)
Keyphrases
- speech recognition
- false positives
- benchmark datasets
- detection method
- decision trees
- feature space
- feature vectors
- low level
- co occurrence
- weak classifiers
- automatic detection
- feature set
- image features
- classification accuracy
- feature extraction
- training data
- information retrieval
- ensemble methods
- classification models
- machine learning