On noise gain estimation for HMM-based speech enhancement

• To address the variation of noise level in non-stationary noise signals, we study the noise gain estimation for speech enhancement using hidden Markov models (HMM). We consider the noise gain as a stochastic process and we approximate the probability density function (PDF) to be log-normal distributed. The PDF parameters are estimated for every signal block using the past noisy signal blocks. The approximated PDF is then used in a Bayesian speech estimator minimizing the Bayes risk for a novel cost function, that allows for an adjustable level of residual noise. As a more computationally efficient alternative, we also derive the maximum likelihood (ML) estimator, assuming the noise gain to be a deterministic parameter. The performance of the proposed gain-adaptive methods are evaluated and compared to two reference methods. The experimental results show significant improvement under noise conditions with time-varying noise energy.

Ämnesord

NATURVETENSKAP  -- Data- och informationsvetenskap (hsv//swe)

NATURAL SCIENCES  -- Computer and Information Sciences (hsv//eng)

Nyckelord

Acoustic noise

Gain measurement

Markov processes

Maximum likelihood estimation

Probability density function

Random processes

Speech recognition

Bayesian speech estimator

Gain-adaptive methods

Hidden Markov models (HMM)

Maximum likelihood (ML)

Acoustic signal processing

Information technology

Informationsteknik

Publikations- och innehållstyp

ref (ämneskategori)

kon (ämneskategori)