oalib
Search Results: 1 - 10 of 100 matches for " "
All listed articles are free for downloading (OA Articles)
Page 1 /100
Display every page Item
About Multichannel Speech Signal Extraction and Separation Techniques  [PDF]
Adel Hidri, Souad Meddeb, Hamid Amiri
Journal of Signal and Information Processing (JSIP) , 2012, DOI: 10.4236/jsip.2012.32032
Abstract: The extraction of a desired speech signal from a noisy environment has become a challenging issue. In the recent years, the scientific community has particularly focused on multichannel techniques which are dealt with in this review. In fact, this study tries to classify these multichannel techniques into three main ones: Beamforming, Independent Component Analysis (ICA) and Time Frequency (T-F) masking. This paper also highlights their advantages and drawbacks. However these previously mentioned techniques could not afford satisfactory results. This fact leads to the idea that a combination of those techniques, which is depicted along this study, may probably provide more efficient results. Indeed, giving the fact that those approaches are still be considered as being not totally efficient, has led us to review these mentioned above in the hope that further researches will provide this domain with suitable innovations.
Research of Blind Source Separation on the Speech Signal Based on Natural Gradient Method  [PDF]
Xinling Wen,Yu Chen
Information Technology Journal , 2012,
Abstract: The Blind Source Separation (BSS) algorithm has got more and more attention in signal processing field. Among the BSS algorithm, the natural gradient algorithm is one of the important method. This study mainly introduces the way of the natural gradient algorithm and realizes an application in the speech signal processing. Through the experiment, when the iteration of 60 times, the natural gradient algorithm on the speech signal separation can realize convergence and the steady-state string sound error approaches to zero which proving the good separation results based on the gradient algorithm on the speech signal separation.
Direct Recovery of Clean Speech Using a Hybrid Noise Suppression Algorithm for Robust Speech Recognition System  [PDF]
Peng Dai,Ing Yann Soon,Rui Tao
ISRN Signal Processing , 2012, DOI: 10.5402/2012/306305
Abstract: A new log-power domain feature enhancement algorithm named NLPS is developed. It consists of two parts, direct solution of nonlinear system model and log-power subtraction. In contrast to other methods, the proposed algorithm does not need prior speech/noise statistical model. Instead, it works by direct solution of the nonlinear function derived from the speech recognition system. Separate steps are utilized to refine the accuracy of estimated cepstrum by log-power subtraction, which is the second part of the proposed algorithm. The proposed algorithm manages to solve the speech probability distribution function (PDF) discontinuity problem caused by traditional spectral subtraction series algorithms. The effectiveness of the proposed filter is extensively compared using the standard database, AURORA2. The results show that significant improvement can be achieved by incorporating the proposed algorithm. The proposed algorithm reaches a recognition rate of over 86% for noisy speech (average from SNR 0?dB to 20?dB), which means a 48% error reduction over the baseline Mel-frequency Cepstral Coefficient (MFCC) system. 1. Introduction The main objective of speech recognition is to get a higher recognition rate. However, lots of factors tend to degrade the performance of automatic speech recognition (ASR) system, such as environmental noise, channel distortion, and speaker variability [1, 2]. Generally, automatic speech recognition system consists of two parts, feature extraction and pattern matching. Therefore, methods which aim to improve the performance of ASR system can be mainly divided into two categories, the “model” approach and the “feature” approach. The “model” approach mainly focuses on improving the speech recognizer, where the speech features are classified into different patterns developed from the statistical properties of speech. As for “feature" approach, emphasis is put on improving the robustness of speech features. The method proposed by this paper belongs to this category. Noise reduction or clean speech estimation is a straight forward “feature” approach to improve the performance of ASR systems. There are different ways to get the estimation. minimum mean square Error is one of the most important ones. Ephraim derived the short-time spectral amplitude (STSA) estimator using minimum mean square error (MMSE) in 1984 [3], which has become a standard approach for clean speech estimation in speech processing. The advantage of MMSE estimator is very obvious. It is mathematically optimized, which theoretically can get a good estimation of the
Fast Noise Compensation and Adaptive Enhancement for Speech Separation  [cached]
Hu Rong,Zhao Yunxin
EURASIP Journal on Audio, Speech, and Music Processing , 2008,
Abstract: We propose a novel approach to improve adaptive decorrelation filtering- (ADF-) based speech source separation in diffuse noise. The effects of noise on system adaptation and separation outputs are handled separately. First, fast noise compensation (NC) is developed for adaptation of separation filters, forcing ADF to focus on source separation; next, output noises are suppressed by speech enhancement. By tracking noise components in output cross-correlation functions, the bias effect of noise on the system adaptation objective function is compensated, and by adaptively estimating output noise autocorrelations, the speech separation output is enhanced. For fast noise compensation, a blockwise fast ADF (FADF) is implemented. Experiments were conducted on real and simulated diffuse noises. Speech mixtures were generated by convolving TIMIT speech sources with acoustic path impulse responses measured in a real room with reverberation time second. The proposed techniques significantly improved separation performance and phone recognition accuracy of ADF outputs.
Fast Noise Compensation and Adaptive Enhancement for Speech Separation  [cached]
Rong Hu,Yunxin Zhao
EURASIP Journal on Audio, Speech, and Music Processing , 2008, DOI: 10.1155/2008/349214
Abstract: We propose a novel approach to improve adaptive decorrelation filtering- (ADF-) based speech source separation in diffuse noise. The effects of noise on system adaptation and separation outputs are handled separately. First, fast noise compensation (NC) is developed for adaptation of separation filters, forcing ADF to focus on source separation; next, output noises are suppressed by speech enhancement. By tracking noise components in output cross-correlation functions, the bias effect of noise on the system adaptation objective function is compensated, and by adaptively estimating output noise autocorrelations, the speech separation output is enhanced. For fast noise compensation, a blockwise fast ADF (FADF) is implemented. Experiments were conducted on real and simulated diffuse noises. Speech mixtures were generated by convolving TIMIT speech sources with acoustic path impulse responses measured in a real room with reverberation time T60=0.3 ¢ € ‰second. The proposed techniques significantly improved separation performance and phone recognition accuracy of ADF outputs.
BSS switch algorithm based on kurtosis and applications in blind separation of speech signal
基于峭度的BSS开关算法的语音信号盲分离*

LIANG Shu-fen,JIANG Tai-hui,
梁淑芬
,江太辉

计算机应用研究 , 2010,
Abstract: The main types of blind signal processing algorithm are batch algorithm and adaptive agorithm.This paper presented the adaptive blind signal separation switch algorithm based on kurtosis for speech signal blind separation processing.Through the comprehensive experiments, the results show that BSS switch algorithm based on kurtosis has good signal separation efficiency from the signal waveforms and spectrums before and after separation and the main evaluation parameters.BSS switch algorithm based on kurtosis has better separation efficiency than the JADE FOBI algorithm.
基于空域信号分离的多径干扰抑制算法
Multipath interference suppression algorithm based on spatial signal separation
 [PDF]

贾琼琼,吴仁彪,王文益,卢丹,王璐
- , 2017,
Abstract: 针对全球导航卫星系统中的多径干扰抑制问题,提出了一种有效的低复杂度多径干扰抑制算法,在直达信号和多径干扰来向角(DOA)未知的情况下,利用RELAX算法,通过反复迭代的思想逐一估计出直达信号和多径干扰的DOA和幅度,进而识别多径干扰,利用线性约束最小方差无畸变波束形成技术,在直达信号方向形成增益与在多径干扰方向形成零陷而抑制多径干扰,并进行了仿真试验。分析结果表明:当多径干扰部分相干时,RELAX算法估计的DOA均方根误差比MUSIC算法低了约12 dB; 当多径干扰完全相干时,RELAX算法估计的DOA均方根误差比MUSIC算法低了约25 dB; 所提算法能够准确估计出直达信号和多径干扰的DOA,保证通过波束形成零陷空域多径干扰,使得进入跟踪环路的信号非常洁净,处理后的码相位跟踪误差接近于0。
Aiming at the multipath interference suppression problem in global navigation satellite system(GNSS), an efficient multipath interference suppression algorithm with low complexity was proposed. Under the unknown directions of arrival(DOAs)of direct signal and multipath interference, the RELAX algorithm was used to estimate the DOAs and amplitudes of direct signal and multipath interference one by one through repeated iteration idea, and then to distinguish multipath interference. The beam forming technology without distortion based on linearly constrained minimum variance(LCMV)was used to form gain at direct signal direction and null at multipath interference direction, and then to suppress multipath interference. The simulation test was carried out. Analysis result shows that when multipath interferences are partially coherent, the root mean square error of DOA estimated by RELAX algorithm is about 12 dB lower than the value estimated by MUSIC algorithm. When multipath interferences are completely coherent, the root mean square error of DOA estimated by RELAX algorithm is about 25 dB lower than the value estimated by MUSIC algorithm. The proposed algorithm can accurately estimate the DOAs of direct signal and multipath interference, which can ensure nulling the multipath interference in airspace through beam forming to make the signal entering the tracking loop very clean. The processed tracking error of code phase is close to 0. 6 figs, 24 refs
DOA Estimation and Signal Recovery Combined Blind Source Separation with High Resolution
盲源分离与高分辨融合的DOA估计与信号恢复方法

KANG Chun-Yu ZHANG Xin-Hua HAN Dong,
康春玉
,章新华,韩东

自动化学报 , 2010,
Abstract: Direction of arrival (DOA) estimation and signal recovery are the base of the underwater target orientation, tracking and recognition, respectively. Based on the array manifold which can be estimated using blind source separation, and by combining the complex blind source separation with the high resolution method, a new method for direction estimation and signal recovery is proposed. It was tested by the simulation with wideband data, the result showed that this method can complete the real-time estimation of the target direction and estimate the corresponding signal of targets. It is superior to the single high resolution method for the same result under the same condition. It was also tested by the recorded data in real sea. Its performance is better than that of the routine minimum variance distortionless response (MVDR) method. It can obviously increase the space spectrum power of the faint target signal and improve the detection capability of the sonar system.
Line Spectral Frequency-based Noise Suppression for Speech-Centric Interface of Smart Devices
JANG, G. J.,PARK, J. S.,KIM, J. H.,SEO, Y. H.
Advances in Electrical and Computer Engineering , 2011, DOI: 10.4316/aece.2011.04001
Abstract: This paper proposes a noise suppression technique for speech-centric interface of various smart devices. The proposed method estimates noise spectral magnitudes from line spectral frequencies (LSFs), using the observation that adjacent LSFs correspond to peak frequencies of spectrum, whereas isolated LSFs are close to flattened valley frequencies retaining noise components. Over a course of segmented time frames, the logarithms of spectral magnitudes at respective LSFs are computed, and their distribution is then modeled by the Rayleigh probability density function. The standard deviation from the Rayleigh function approximates the noise spectral magnitude. The model is updated at every frame in an online manner so that it can deal with real-time inputs. Once the noise spectral magnitude is estimated, a time-domain Wiener filter is derived for the suppression of the estimated noise spectral magnitude, and this is then applied to the input noisy speech signals. Our proposed approach operates well on most smart devices owing to its low computational complexity and real-time implementation. Speech recognition experiments, conducted to evaluate the proposed technique, show that our method exhibits superior performance, with less distortion of original speech, when compared to conventional noise suppression techniques.
Speech Signal Noise Reduction by Wavelets
Roopali Goel,Ritesh Jain
International Journal of Innovative Technology and Exploring Engineering , 2013,
Abstract: Speech plays an important role in multimedia system. Speech enhancement is to remove noise from speech for multimedia systems. Noise act as a disturbance in any form of communication which degrades the quality of the information signal. Generally transmission and receiving signals are often corrupted by noise which can cause severe problems for downstream processing and user perception. Therefore an automated removal of noise would be an invaluable first stage for many signal processing tasks. Denoising has long been a focus of research and yet there always remains room for improvement. There are so many ways to improve the signal quality or to regenerate the signal. In this paper we have present a method for speech signal denoising using different wavelets. In this we will demonstrate the usefulness of wavelets to reduce noise in a model system where Gaussian noise is inserted into an audio signal.
Page 1 /100
Display every page Item


Home
Copyright © 2008-2017 Open Access Library. All rights reserved.