The GMM-SVM Supervector Approach for the Recognition of the Emotional Status from Speech.
Friedhelm SchwenkerStefan SchererYasmine M. MagdiGünther PalmPublished in: ICANN (1) (2009)
Keyphrases
- gaussian mixture model
- speaker recognition
- speaker identification
- mel frequency cepstral coefficients
- feature vectors
- recognition engine
- digit recognition
- support vector machine svm
- recognition rate
- support vector machine
- feature space
- speaker independent
- emotional state
- object recognition
- support vector
- mixture model
- feature extraction
- noisy environments
- pattern recognition
- emotion recognition
- recognition accuracy
- automatic speech recognition systems
- em algorithm
- voice activity detection
- speech corpus
- knn
- gaussian mixture
- speech signal
- spoken words
- speech recognition
- continuous speech recognition
- kernel function
- human activities
- background subtraction
- automatic transcription
- spoken language
- feature selection
- training data
- model selection
- expectation maximization
- maximum likelihood
- speech recognition systems
- training set
- text recognition
- moving objects
- speaker verification
- svm classifier
- classification algorithm
- multi class
- text classification