N-Best Tokenization in a GMM-SVM Language Identification System.
Xi YangMan-Hung SiuPublished in: ICASSP (4) (2007)
Keyphrases
- language identification
- speaker identification
- gaussian mixture model
- support vector
- feature vectors
- speech recognition
- knn
- feature selection
- feature extraction
- speech signal
- noisy environments
- mixture model
- named entities
- broadcast news
- document images
- machine learning
- training set
- feature space
- speaker verification
- em algorithm
- training data
- web information retrieval
- expectation maximization
- maximum likelihood
- pattern recognition