Speech Enhancement Based on MAP Estimation Using a Variable Speech Distribution.
Yuta TsukamotoArata KawamuraYouji IiguniPublished in: IEICE Trans. Fundam. Electron. Commun. Comput. Sci. (2007)
Keyphrases
- speech enhancement
- map estimation
- speech signal
- noisy environments
- maximum a posteriori
- noise reduction
- signal to noise ratio
- vocal tract
- single channel
- markov random field
- linear prediction
- maximum likelihood
- noisy images
- smoothing algorithm
- noisy speech
- background noise
- speech recognition
- additive noise
- em algorithm
- probability distribution
- wiener filter
- multi channel
- image reconstruction
- energy function
- automatic speech recognition
- frequency domain
- non stationary
- expectation maximization
- hidden markov models
- pairwise