Robust speaker verification in colored noise environment

International Journal of Speech Technology, 2004

Investigating Speaker Verification in real-world noisy environments, a novel feature extraction process suitable for suppression of time-varying noise is compared with a fine-tuned spectral subtraction method. The proposed feature extraction process is based on approximating the clean speech and the noise spectral magnitude with a mixture of Gaussian probability density functions (pdfs) by using the Expectation-Maximization algorithm (EM). Subsequently, the Bayesian inference framework is applied to the degraded spectral coefficients, and by employing Minimum Mean Square Error Estimation (MMSE), a closed form solution for the spectral magnitude estimation task is derived. The estimated spectral magnitude finally is incorporated into the Mel-Frequency Cepstral Coefficients (MFCCs) front-end of a baseline text-independent speaker verification system, based on Probabilistic Neural Networks, which participated successfully in the 2002 NIST (National Institute of Standards and Technology of USA) Speaker Recognition Evaluation. A comparative study of the proposed technique for real-world noise types demonstrates a significant performance gain compared to the baseline speech features and to the spectral subtraction enhancement method.

International Journal On …, 2011

Abstract: There are many situations where speech is affected by different kind of acoustic noise. We propose an improved spectral subtraction method for the reduction of colored acoustic noise added to the speech. Our implementation uses a multi-band spectral over ...

IASET, 2013

In this paper we study the performance of speaker verification system by applying spectral subtraction to multiband speech in real environment. In real world environment noise from different sources could exist and which may interfere with the speech signal at different frequencies. Because of the colored nature of noise which does not spread uniformly over the spectrum of speech i.e. some of the frequencies may be most affected while some frequencies may be least affected, a multiband filter bank approach is proposed. In this approach a filter bank is designed which divides the speech signal into a number of frequency bands. Spectral subtraction is then applied to each of the bands and the result of all the subtraction are combined at the end. Results have shown quite a significant improvement in performance when spectral subtraction is applied to multi band than applied to the entire speech signal.

ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings, 2012

Different short-term spectrum estimators for speaker verification under additive noise are considered. Conventionally, mel-frequency cepstral coefficients (MFCCs) are computed from discrete Fourier transform (DFT) spectra of windowed speech frames. Recently, linear prediction (LP) and its temporally weighted variants have been substituted as the spectrum analysis method in speech and speaker recognition. In this paper, 12 different short-term spectrum estimation methods are compared for speaker verification under additive noise contamination. Experimental results conducted on NIST 2002 SRE show that the spectrum estimation method has a large effect on recognition performance and stabilized weighted LP (SWLP) and minimum variance distortionless response (MVDR) methods yield approximately 7 % and 8 % relative improvements over the standard DFT method at -10 dB SNR level of factory and babble noises, respectively in terms of equal error rate (EER).

2012

Speech enhancement aims to improve speech quality by using various techniques and algorithms. Over the past several years there has been considerable attention focused on the problem of enhancement of speech degraded by additive background noise. Background noise suppression has many applications. Using mobile in a noisy environment like in streets or in a car is an obvious application, removing the background noise when sending speech from the cockpit of an airplane to the ground or to the cabin. The spectral subtractive algorithm is historically one of the first algorithms proposed for additive background noise and it has gone through many modifications with time. This is a review paper and its objective is to provide an overview of the variety of spectral subtraction techniques that have been proposed for enhancement of speech degraded by additive background noise during past decades . Section I gives the Introduction to Speech enhancement and explain basic Spectral Subtraction t...

2005

Classical speech enhancement techniques and recently developed wavelet denoising schemes are applied to speaker verification systems in noise. Merely applying these techniques to corrupted testing signals does not properly decreases the error rates when clean speech is used for training signals. In this paper, a noise modelling approach is used to corrupt the training signals according to an estimate of the noise present in the test signal. We show that this procedure makes the error rates drop to a fraction of the original results.

Interspeech 2010, 2010

It is well known that MFCC based speaker identification (SID) systems easily break down under mismatched training and test conditions. In this paper, we report on a study that considers four different single-channel speech enhancement front-ends for robust SID under such conditions. Speech files from the YOHO database are corrupted with four types of noise including babble, car, factory, and white Gaussian at five SNR levels (0-20 dB), and processed using four speech enhancement techniques representing distinct classes of algorithms: spectral subtraction, statistical model-based, subspace, and Wiener filtering. Both processed and unprocessed files are submitted to a SID system trained on clean data. In addition, a new set of acoustic feature parameters based on Hilbert envelope of gammatone filterbank outputs are proposed and evaluated for SID task. Experimental results indicate that: (i) depending on the noise type and SNR level, the enhancement front-ends may help or hurt SID performance, (ii) the proposed feature significantly achieves higher SID accuracy compared to MFCCs under mismatched conditions.

2011

Speech enhancement aims to improve speech quality by using various techniques and algorithms. The Spectral subtraction technique is historically one of the first algorithms proposed for removal of additive background noise. It is a single channel speech enhancement technique for the enhancement of speech degraded by additive background noise. Background noise can effect our conversation in a noisy environment like in streets or in a car, when sending speech from the cockpit of an airplane to the ground or to the cabin and can effect both quality and intelligibility of speech. With the passage of time Spectral subtraction has undergone many modifications. This is a review paper and its objective is to provide an overview of the variety of spectral subtraction techniques that have been proposed for enhancement of speech degraded by additive background noise during past decades . Section I gives the Introduction to Speech enhancement and explain basic Spectral Subtraction technique. Se...