Forensic speaker verification using formant features and Gaussian mixture models.
Timo BeckerMichael JessenCatalin GrigorasPublished in: INTERSPEECH (2008)
Keyphrases
- speaker recognition
- gaussian mixture model
- speaker verification
- mel frequency cepstral coefficients
- feature vectors
- speech signal
- mixture model
- feature space
- feature extraction
- extracting features
- feature set
- speech recognition
- acoustic features
- noisy environments
- speaker identification
- neural network
- vector quantization
- maximum likelihood
- em algorithm
- image segmentation
- learning algorithm
- image processing
- multimedia
- using artificial neural networks
- pattern recognition