Low-Frequency Bandwidth Extension of Telephone Speech Using Sinusoidal Synthesis and Gaussian Mixture Model.
Hannu PulakkaUlpu RemesSanteri YrttiahoKalle J. PalomäkiMikko KurimoPaavo AlkuPublished in: INTERSPEECH (2011)
Keyphrases
- gaussian mixture model
- low frequency
- high frequency
- speaker recognition
- speaker identification
- mixture model
- feature vectors
- visual quality
- subband
- high resolution
- maximum likelihood
- discrete wavelet transform
- frequency domain
- em algorithm
- feature space
- wavelet domain
- expectation maximization
- wavelet transform
- speech recognition
- wavelet coefficients
- gaussian mixture modeling
- background subtraction
- video streaming
- generative model
- frequency band
- facial animation
- image compression
- object recognition
- speech signal
- feature selection
- computer vision
- information retrieval