Speech Enhancement Based on Compressed Sensing Technology
نویسنده
چکیده
Compressed sensing (CS) is a sampled approach on signal sparsity-base, and it can effectively extract the information which is contained in the signal. This paper presents a noisy speech enhancement new method based on CS process. Algorithm uses a voice sparsity in the discrete fast Fourier transform (Fast Fourier transform, FFT), and complex domain observation matrix is designed, and the noisy speech compression measurement and de-noising are made by soft threshold, and the speech signal is sparsely reconstructed by separable approximation (Sparse Reconstruction by Separable Approximation, SpaRSA) algorithm to restore, speech enhancement is improved. Experimental results show that the denoising compression reconstruction of the noisy signal is done in the algorithm, SNR margin is improved greatly, and the background noise can been more effectively suppressed. Copyright © 2014 IFSA Publishing, S. L.
منابع مشابه
Noisy Signal Processing Research based on Compressed Sensing Technology
Compressed sensing (CS) is a kind of sampling method based on signal sparse property, it can effectively extract the signal which was contained in the message. In this study, a new noise speech enhancement method was proposed based on CS process. Voice sparsity is used to this algorithm in the discrete fast Fourier transform (Fast Fourier transform, FFT), and observation matrix is designed in c...
متن کاملSpeech enhancement using compressed sensing
This paper proposes an approach based on the compressed sensing for speech enhancement. It attempts to exploit the fact that it is relatively easy to find sparse representation for speech signal but the same cannot hold true for noise. Thus for a given speech signal a sparse vector is derived which extracts the speech signal from the given noisy speech signal. Approach is further modified to en...
متن کاملStudy on Joint Speech Encoding Technology based on Compressed Sensing
In this paper, a new joint speech encoding scheme based on compressed sensing was proposed. In this encoding algorithms, the compressed sensing reconstruction and PCM (Pulse Coding Modulation) were used for the speech signal encode. For the speech signal, the high frequency and low frequency coefficients can be acquired by using the wavelet transform based on the lifting scheme. For the details...
متن کاملSpeech enhancement based on hidden Markov model using sparse code shrinkage
This paper presents a new hidden Markov model-based (HMM-based) speech enhancement framework based on the independent component analysis (ICA). We propose analytical procedures for training clean speech and noise models by the Baum re-estimation algorithm and present a Maximum a posterior (MAP) estimator based on Laplace-Gaussian (for clean speech and noise respectively) combination in the HMM ...
متن کاملSpeech Enhancement based on Wiener Filter and Compressive Sensing
In the last few decades, many advanced technologies have been proposed, in which communications played a great role as well as telecommunications applications. The noise elimination in various environments became the most concerned as it greatly hindered the speech communication applications. The improvement of noisy speech interms of quality and intelligibility are taken into account without i...
متن کامل