Noise spectrum estimation is a fundamental component of speech enhancement and
speech recognition systems. We present an improved minima controlled recursive …

A noise-estimation algorithm is proposed for highly non-stationary noise environments. The
noise estimate is updated by averaging the noisy speech power spectrum using time and …

How to construct models for speech/nonspeech discrimination is a crucial point for voice
activity detectors (VADs). Semi-supervised learning is the most popular way for model …

A single-microphone noise suppression algorithm is described that is based on a novel
approach for the estimation of the signal-to-noise ratio (SNR) in different frequency …

Automatic speech recognition (ASR) has reached very high levels of performance in
controlled situations. However, the performance degrades significantly when environmental …

This paper presents a novel speech feature enhancement technique based on a
probabilistic, nonlinear acoustic environment model that effectively incorporates the phase …

Although the field of automatic speaker recognition (ASR) has been the subject of extensive
research over the past decades, the lack of robustness against background noise has …

In this paper, the performance evaluation of Modified Cascaded Median (MCM)-based noise
estimation method for speech enhancement system has been carried out. The MCM-based …

In this chapter, we focus on the statistical methods that constitute a speech spectral
enhancement system and describe some of their fundamental components. We begin in …

Real world applications such as hands-free dialling in cars may have to deal with potentially
very noisy environments. Existing state-of-the-art solutions to this problem use feature-based …