GMM/SVM N-best speaker identification under mismatch channel conditions.
Ilija ZeljkovicPatrick HaffnerBrian AmentoJay G. WilponPublished in: ICASSP (2008)
Keyphrases
- speaker identification
- gaussian mixture model
- channel conditions
- support vector
- speaker recognition
- speech recognition
- feature vectors
- speech signal
- noisy environments
- rate adaptation
- feature extraction
- channel coding
- packet loss
- broadcast news
- knn
- wireless channels
- feature space
- source coding
- video transmission
- bitstream
- bit rate
- maximum likelihood
- feature selection
- machine learning
- rate distortion
- mixture model
- training set
- coding scheme
- neural network