Speech enhancement using minimum mean-square error estimation and a post-filter derived from vector quantization of clean speech.
Jason WungShigeki MiyabeBiing-Hwang JuangPublished in: ICASSP (2009)
Keyphrases
- vector quantization
- speech enhancement
- error estimation
- noise reduction
- noisy speech
- noisy environments
- speech signal
- wiener filter
- image compression
- signal to noise ratio
- speech recognition
- single channel
- vector quantizer
- smoothing algorithm
- vocal tract
- model selection
- linear prediction
- speaker recognition
- edge detection
- hidden markov models
- background noise
- speech synthesis
- multi channel
- image coding
- support vector machine
- generalization error
- speaker verification
- non stationary
- language model
- denoising
- pairwise