Multimodal speaker identification using an adaptive classifier cascade based on modality reliability.
Engin ErzinYücel YemezA. Murat TekalpPublished in: IEEE Trans. Multim. (2005)
Keyphrases
- speaker identification
- probabilistic neural network
- multi modal
- speech recognition
- gaussian mixture model
- feature transformation
- noisy environments
- speech signal
- feature extraction
- training data
- broadcast news
- linear classifiers
- feature space
- training set
- weak classifiers
- support vector
- feature selection
- multimedia
- automatic speech recognition
- machine learning
- noise reduction
- high level
- language model
- feature set
- natural language processing