Speech Enhancement in Noisy Environments in Hearing Aids Driven by a Tailored Gain Function Based on a Gaussian Mixture Model.
Lorena ÁlvarezEnrique AlexandreCosme LlerenaRoberto Gil-PitaLucas CuadraPublished in: ICAISC (1) (2013)
Keyphrases
- speech enhancement
- gaussian mixture model
- noisy environments
- hearing aids
- noise reduction
- speaker identification
- speech signal
- signal to noise ratio
- speaker recognition
- mixture model
- speech recognition
- speaker verification
- single channel
- feature vectors
- spectral subtraction
- em algorithm
- background noise
- linear prediction
- maximum likelihood
- expectation maximization
- vocal tract
- wiener filter
- gaussian distribution
- noisy speech
- hidden markov models
- additive noise
- smoothing algorithm
- language model
- automatic speech recognition
- image restoration
- generative model