Bandwidth Extension of Telephone Speech to Low Frequencies Using Sinusoidal Synthesis and a Gaussian Mixture Model.
Hannu PulakkaUlpu RemesSanteri YrttiahoKalle J. PalomäkiMikko KurimoPaavo AlkuPublished in: IEEE Trans. Speech Audio Process. (2012)
Keyphrases
- gaussian mixture model
- speaker recognition
- speaker identification
- mixture model
- feature vectors
- background subtraction
- probabilistic neural network
- maximum likelihood
- expectation maximization
- em algorithm
- speech recognition
- gaussian mixture modeling
- gaussian model
- gaussian mixture
- gaussian distribution
- speech signal
- computer vision
- prior knowledge
- pairwise
- noisy environments
- feature space
- pattern recognition
- video sequences
- bayesian information criterion
- machine learning