Low rate coders based on the harmonic-noise model are sensitive to acoustic background noise at low SNRs due to the increase in parameter errors from the analysis of noisy speech. We investigate the use of spectral subtraction enhancement preprocessing on the performance of the sinusoidal model based codec both by objective assessment of parameter errors and the subjective testing of output spe...

In this paper, different feature extraction methods for speech recognition system such as Melfrequency cepstral coefficients (MFCC), linear predictive coefficient cepstrum (LPCC) and Bark frequency cepstral coefficients (BFCC) are implemented and the comparison is done based on average recognition accuracy. We suggest a noise robust isolated word speech recognition system which can be applied i...

In this paper, a speech enhancement method based on noise compensation performed on short time magnitude as well phase spectra is presented. Unlike the conventional geometric approach (GA) to spectral subtraction (SS), here the noise estimate to be subtracted from the noisy speech spectrum is proposed to be determined by exploiting the low frequency regions of current frame of noisy speech rath...

Two noisy speech processing problems—speech enhancement and noisy speech recognition—are dealt with in this paper. The technique we focus on is by using the filtering approach; a novel filter, the recurrently adaptive fuzzy filter (RAFF), is proposed and applied to these two problems. The speech enhancement is based on adaptive noise cancellation with two microphones, where the RAFF is used to ...

In this paper we propose an architecture for low bit rate coding of noisy speech. The input noisy speech is decomposed into multi-resolution signal components using wavelet transform. An iterative Wiener lter-ing is used at each level of wavelet analysis to enhance speech. The system model that evolves during enhancement is processed further to get optimal parameters for the quantization. A mul...

A new, single-ended, i.e. reference-free measure for the prediction of perceived listening effort of noisy speech is presented. It is based on phoneme posterior probabilities (or posteriorgrams) obtained from a deep neural network of an automatic speech recognition system. Additive noisy or other distortions of speech tend to smear the posteriorgrams. The smearing is quantified by a performance...

Recently, speech recognition performance has been drastically improved by statistical methods and huge speech databases. Now performance improvement under such realistic environments as noisy conditions is being focused on. Since October 2001, we from the working group of the Information Processing Society in Japan have been working on evaluation methodologies and frameworks for Japanese noisy ...

The classification of active speech vs. inactive speech in noisy speech is an important part of speech applications, typically in order to achieve a lower bit-rate. In this work, the error rates for raw classification (i.e. with no hangover mechanism) of noisy speech obtained with traditional classification algorithms are compared to the rates obtained with Neural Network classifiers, trained w...