Laguerre滤波器在语音识别前端处理中的应用研究
|
摘要
语音识别是语音信号处理领域的研究热点，但由于其研究的复杂性，长期以来一直是一项难题，尤其是噪声环境下的非特定人语音识别。本文从一个典型的语音识别系统出发，介绍了语音识别的基本原理，讨论了几种常用的特征提取方法，尤其对过零率峰值幅度(ZCPA)特征提取作了较为详细的介绍。在此基础上提出用Laguerre滤波器对ZCPA特征提取前端处理进行改进的方法，并获得了具有优良抗噪性的识别结果。
本文中用Laguerre网络实现的滤波器吸收了传统有限冲激响应(FIR)、无限冲激响应(IIR)滤波器的优点，既具有FIR滤波器的稳定性又具有IIR滤波器的长时记忆的特点和通阻带特性。其设计方法是在Laguerre滤波器与理想滤波器的频率响应的均方误差为最小的前提下，利用牛顿-拉夫逊法估算滤波器参数，然后由柯西-留数定理得出相对应的Laguerre系数以获得最优滤波器。通过实例设计了Laguerre滤波器，并与传统FIR和IIR滤波器的频率响应作了详细的比较，得出Laguerre滤波器有较小的滤波器长度，合适的线性相位和较少的通阻带波纹。缺点是计算复杂，但使用介绍的引理可降低其复杂性。
接着将Laguerre滤波器用在ZCPA特征提取中代替原来的FIR滤波器，后端分别利用RBF网络和HMM训练和识别。实验结果表明利用Laguerre滤波器代替FIR滤波器进行特征提取，其识别率明显提高，而且抗噪性有很大改善。
论文最后分析了Laguerre序列的频率弯折特性，并将小波变换的多分辨特性与之相结合得出基于Laguerre网络的频率弯折小波变换，对其实现结构作了详纽介绍，同时也说明难点所在。提出下一步工作是将频率弯折小波变换用于特征提取中，期望得到好的识别结果。
Speech recognition has become a hotspot in the field of speech signal processing. But it is not easy to solve perfectly because of its complexity, especially for the speech recognition of Speaker Independent in noisy environment. This paper introduced the fundamental of speech recognition and discussed some commonly used feature extraction methods and specially analysed the ZCPA feature extraction based on a classical speech recognition system. Based on above conclusion, it presented a method which improved the front-end processing of ZCPA feature extraction, and got the better recognition rate which has excellent anti-noise properties.
The filter realized by Laguerre network is a compromise between the FIR and IIR. It not only possessed the stability of FIR, but also had the good property of pass-band and stop-band of IIR, and achieved a long time memory. In this paper, Laguerre filter's design approach was performed by evaluating the filter parameter employed Newton-Raphson method and corresponding Laguerre coefficients and obtaining optimum filter employed Canchy theorem when the minimum-mean-square-error of the frequency response between the Laguerre filter and the optimum filter was existed. It designed Laguerre filter and compared the frequency response of Laguerre filter with the frequency response of conventional FIR and IIR filters through experiment, educed Laguerre filter has small length and appropriate linear phase with less ripples in pass-band and stop-band. But the cost is the complex computation which can be reduced by the introduced lemma.
Based on the better property of Laguerre filter, it used Laguerre filter replacing the traditional FIR filter in the ZCPA feature extraction, and employed RBF network and HMM to train and recognise in the back-end. The experiment results showed applying Laguerre filter replacing the traditional FIR filter in feature extraction would improved speech recognition rate and anti-noise properties.
At last, it combined Laguerre transform and wavelet transform based on the frequency warped properties of Laguerre transform and multiresolution of wavelet transform, detailedly introduced its structure, simultaneously also explained what is the difficulty, and the further task is applying the frequency warped wavelet transform to the feature extraction, it's expect to gain the better speech recognition result.
本文中用Laguerre网络实现的滤波器吸收了传统有限冲激响应(FIR)、无限冲激响应(IIR)滤波器的优点，既具有FIR滤波器的稳定性又具有IIR滤波器的长时记忆的特点和通阻带特性。其设计方法是在Laguerre滤波器与理想滤波器的频率响应的均方误差为最小的前提下，利用牛顿-拉夫逊法估算滤波器参数，然后由柯西-留数定理得出相对应的Laguerre系数以获得最优滤波器。通过实例设计了Laguerre滤波器，并与传统FIR和IIR滤波器的频率响应作了详细的比较，得出Laguerre滤波器有较小的滤波器长度，合适的线性相位和较少的通阻带波纹。缺点是计算复杂，但使用介绍的引理可降低其复杂性。
接着将Laguerre滤波器用在ZCPA特征提取中代替原来的FIR滤波器，后端分别利用RBF网络和HMM训练和识别。实验结果表明利用Laguerre滤波器代替FIR滤波器进行特征提取，其识别率明显提高，而且抗噪性有很大改善。
论文最后分析了Laguerre序列的频率弯折特性，并将小波变换的多分辨特性与之相结合得出基于Laguerre网络的频率弯折小波变换，对其实现结构作了详纽介绍，同时也说明难点所在。提出下一步工作是将频率弯折小波变换用于特征提取中，期望得到好的识别结果。
Speech recognition has become a hotspot in the field of speech signal processing. But it is not easy to solve perfectly because of its complexity, especially for the speech recognition of Speaker Independent in noisy environment. This paper introduced the fundamental of speech recognition and discussed some commonly used feature extraction methods and specially analysed the ZCPA feature extraction based on a classical speech recognition system. Based on above conclusion, it presented a method which improved the front-end processing of ZCPA feature extraction, and got the better recognition rate which has excellent anti-noise properties.
The filter realized by Laguerre network is a compromise between the FIR and IIR. It not only possessed the stability of FIR, but also had the good property of pass-band and stop-band of IIR, and achieved a long time memory. In this paper, Laguerre filter's design approach was performed by evaluating the filter parameter employed Newton-Raphson method and corresponding Laguerre coefficients and obtaining optimum filter employed Canchy theorem when the minimum-mean-square-error of the frequency response between the Laguerre filter and the optimum filter was existed. It designed Laguerre filter and compared the frequency response of Laguerre filter with the frequency response of conventional FIR and IIR filters through experiment, educed Laguerre filter has small length and appropriate linear phase with less ripples in pass-band and stop-band. But the cost is the complex computation which can be reduced by the introduced lemma.
Based on the better property of Laguerre filter, it used Laguerre filter replacing the traditional FIR filter in the ZCPA feature extraction, and employed RBF network and HMM to train and recognise in the back-end. The experiment results showed applying Laguerre filter replacing the traditional FIR filter in feature extraction would improved speech recognition rate and anti-noise properties.
At last, it combined Laguerre transform and wavelet transform based on the frequency warped properties of Laguerre transform and multiresolution of wavelet transform, detailedly introduced its structure, simultaneously also explained what is the difficulty, and the further task is applying the frequency warped wavelet transform to the feature extraction, it's expect to gain the better speech recognition result.
引文
[01] 何湘智，语音识别的研究与发展，计算机与现代化，2002，(3)，3-6
[02] 俞铁城，语音识别的发展现状，通讯世界，2005，(2)，56-57
[03] Rabiner L, Toward Vision, Voice and Audio Processing Consideration, AT&T Technical, 1997, 2(4), 4-13
[04] 江铭虎，袁保宗，林碧琴，神经网络语音识别的研究及进展，电信科学，1997，13(7)，1-5
[05] 拉宾纳，语音识别的基本原理，北京，清华大学出版社，2002
[06] 赵力著，语音信号处理，北京，机械工业出版社，2003
[07] 胡浩松，滤波器的过去现在和未来，http://eelab.sjtu.edu.cn/dg/xsyd/011.doc
[08] 杨行峻，迟惠生等，语音信号处理，北京，电子工业出版社，1995
[09] 胡航，语音信号处理，黑龙江，哈尔滨工业大学出版社，2000
[10] B.A. Dautrich, L.R.Rabiner, T.B.Martin, "On the Effects of Varying Filter Bank Parameters on Isolated Word Recognition," IEEE Trans, Acoustics, Speech, Signal Proc., 1983, 31(4), 793-807
[11] 于秀丽，对神经网络学习算法的研究，[硕士学位论文]，河北，河北工业大学，2003
[12] 蔡莲红，黄德智，蔡锐，现代语音技术基础与应用，北京，清华大学出版社，2003
[13] 李鸣华，一种基于听觉模型的语音特征提取方法，计算机与现代化，2000，第3期，9-13
[14] Doh-Suk Kim, Soo-Yong Lee, Rhee M.Kil, "Auditory Processing of Speech Signals for Robust Speech Recognition in Real World Noisy Environments", IEEE Transactions on Speech and Audio Processing, 1999, 7(1), 1063-1067
[15] 梁凤岗，FIR数字滤波器设计的小波逼近方法，上海交通大学学报，2004，35(1)
[16] 杨永昌，李晨辉，FIR数字滤波器的设计方法，桂林航天工业高等专科学校校报，2006，43(3)
[17] Arojit Roychowdhury, "FIR Filter Design Techniques", Submitted to IEEE, 2002
[18] Sanjit K．Mitra(著)，孙洪，余翔宇等(译)，数字信号处理(基于计算机的方法)，北京，电子工业出版社，2005
[19] 贺双赤，用Laguerre滤波器实现多径衰落信道自适应均衡，电讯技术，2004，44(1)，82-86
[20] Masnadi-Shirazi MA, Aleshams M, "Laguerre Discrete-time Filter Design", Computers and Electrical Engineering, 2003, (29), 173-192
[21] Amole CR, "Laguerre Function and the Laguerre Nerwork-Their Properties and Digital Simulation", Massachusetts Institute of Technology Lincoln Laboratory, Technical Note, 1966, 1966-1978
[22] King RE, Pareskevopoulos PN, "Digital Laguerre Filters", Circ Theory Appl, 1977, (5), 81-91
[23] Masnadi-Shirazi MA, Ahmed N, "Optimum Laguerre Networks for a Class of Discrete-time Systems", IEEE Trans.on Signal Processing, 1991, 39(9), 2104-2108
[24] Silva Toe, "On the Determination of the Optimal Pole Position of Laguerre Filters", IEEE Trans Signal Processing, 1995, 4(9), 2079-2087
[25] Masnadi-Shirazi MA, Ghasemi M, "Laguerre Digital Filter Design", Proc.IEEE Int.Conf. Acoust, Speech and Dignal Processing, 1 (2), Detroit, 1995, 1284-1287
[26] Wylie CR, Barret LC, "Advanced Engineering Mathematics", New York, McGraw-Hill, 1966, 695[chapter 15]
[27] Ahmed N, Natarajan T, "Disctete-time Signals and Systems", Reston Publishing Company, Reston, 1983, 137[chapter 4]
[28] Oppenheim AV, Schafer RW, "Discrete-time Signal Processing", Englewood Cliffs, NJ: Prentice-Hall, 1989[chapter 7]
[29] 吴玺宏，迟惠生，王楚，基于听觉外周模型的语音信号听觉神经表示，生物物理学报，1997，13(2)，213-220
[30] 张刚，张雪英，语音处理与编码，北京，兵器工业出版社，2000
[31] Oded Ghitza, "Auditory Models and Human Performance in Tasks Related to Speech Coding and Speech Recognition", IEEE Transactions on Speech and Audio Processing, 1994, 2(1), 113-131
[32] 焦志平，改进的ZCPA语音特征提取算法研究，[硕士学位论文]，太原，太原理工大学，2005
[33] Siu-Yeung Cho, Tommy W.S.Chow, "Learning Parametric Specular Reflectance Model by Radial Basis Function Network", IEEE Transactions on Neural Networks, 2000, 11 (6), 1498-1503
[34] Martin Hagan, Howard B. Demuth, Mark H. Beale, "Neural Network Design"，北京，机械工业出版社，2002，197-207
[35] 陈逢时，子波变换理论及其在信号处理中的应用，北京，国防工业出版社，1998
[36] G.Ecangelista, S.Cavaliere, "Frequency Warped Filter Banks and Wavelet Transforms: A Discrete-Time Approach Via Laguerre Expansion," submitted to IEEE Trans. on Signal Processing, 1996
[37] G.Ecangelista, S.Cavaliere, "Discrete Frequency Warped WaveletS: Theory and Application," submitted to IEEE Trans.on Signal Processing, special issue on Theory and Application of Filter banks and Wavelets, 1997
[38] G.Ecangelista, S.Cavaliere, "The Discrete-Time Frequency Warped Wavelet Transforms," Proc.of ICASSP'97, Munich, 1997, 2105-2108
[39] Alan V.Oppenheim, Alan S. Willsky(著)，刘树棠(译)，信号与系统，2，陕西，西安交通大学出版社，1998
[40] 王学辉，张明辉，Matlab6.1最新应用详解，北京，中国水利水电出版社，2002
[41] 薛定宇，陈阳泉，高等应用数学问题Matlab求解，北京，清华大学出版社，2004
[42] 赵高峰，基于小波分析的语音端点检测算法研究，[硕士学位论文]，太原，太原理工大学，2006
[02] 俞铁城，语音识别的发展现状，通讯世界，2005，(2)，56-57
[03] Rabiner L, Toward Vision, Voice and Audio Processing Consideration, AT&T Technical, 1997, 2(4), 4-13
[04] 江铭虎，袁保宗，林碧琴，神经网络语音识别的研究及进展，电信科学，1997，13(7)，1-5
[05] 拉宾纳，语音识别的基本原理，北京，清华大学出版社，2002
[06] 赵力著，语音信号处理，北京，机械工业出版社，2003
[07] 胡浩松，滤波器的过去现在和未来，http://eelab.sjtu.edu.cn/dg/xsyd/011.doc
[08] 杨行峻，迟惠生等，语音信号处理，北京，电子工业出版社，1995
[09] 胡航，语音信号处理，黑龙江，哈尔滨工业大学出版社，2000
[10] B.A. Dautrich, L.R.Rabiner, T.B.Martin, "On the Effects of Varying Filter Bank Parameters on Isolated Word Recognition," IEEE Trans, Acoustics, Speech, Signal Proc., 1983, 31(4), 793-807
[11] 于秀丽，对神经网络学习算法的研究，[硕士学位论文]，河北，河北工业大学，2003
[12] 蔡莲红，黄德智，蔡锐，现代语音技术基础与应用，北京，清华大学出版社，2003
[13] 李鸣华，一种基于听觉模型的语音特征提取方法，计算机与现代化，2000，第3期，9-13
[14] Doh-Suk Kim, Soo-Yong Lee, Rhee M.Kil, "Auditory Processing of Speech Signals for Robust Speech Recognition in Real World Noisy Environments", IEEE Transactions on Speech and Audio Processing, 1999, 7(1), 1063-1067
[15] 梁凤岗，FIR数字滤波器设计的小波逼近方法，上海交通大学学报，2004，35(1)
[16] 杨永昌，李晨辉，FIR数字滤波器的设计方法，桂林航天工业高等专科学校校报，2006，43(3)
[17] Arojit Roychowdhury, "FIR Filter Design Techniques", Submitted to IEEE, 2002
[18] Sanjit K．Mitra(著)，孙洪，余翔宇等(译)，数字信号处理(基于计算机的方法)，北京，电子工业出版社，2005
[19] 贺双赤，用Laguerre滤波器实现多径衰落信道自适应均衡，电讯技术，2004，44(1)，82-86
[20] Masnadi-Shirazi MA, Aleshams M, "Laguerre Discrete-time Filter Design", Computers and Electrical Engineering, 2003, (29), 173-192
[21] Amole CR, "Laguerre Function and the Laguerre Nerwork-Their Properties and Digital Simulation", Massachusetts Institute of Technology Lincoln Laboratory, Technical Note, 1966, 1966-1978
[22] King RE, Pareskevopoulos PN, "Digital Laguerre Filters", Circ Theory Appl, 1977, (5), 81-91
[23] Masnadi-Shirazi MA, Ahmed N, "Optimum Laguerre Networks for a Class of Discrete-time Systems", IEEE Trans.on Signal Processing, 1991, 39(9), 2104-2108
[24] Silva Toe, "On the Determination of the Optimal Pole Position of Laguerre Filters", IEEE Trans Signal Processing, 1995, 4(9), 2079-2087
[25] Masnadi-Shirazi MA, Ghasemi M, "Laguerre Digital Filter Design", Proc.IEEE Int.Conf. Acoust, Speech and Dignal Processing, 1 (2), Detroit, 1995, 1284-1287
[26] Wylie CR, Barret LC, "Advanced Engineering Mathematics", New York, McGraw-Hill, 1966, 695[chapter 15]
[27] Ahmed N, Natarajan T, "Disctete-time Signals and Systems", Reston Publishing Company, Reston, 1983, 137[chapter 4]
[28] Oppenheim AV, Schafer RW, "Discrete-time Signal Processing", Englewood Cliffs, NJ: Prentice-Hall, 1989[chapter 7]
[29] 吴玺宏，迟惠生，王楚，基于听觉外周模型的语音信号听觉神经表示，生物物理学报，1997，13(2)，213-220
[30] 张刚，张雪英，语音处理与编码，北京，兵器工业出版社，2000
[31] Oded Ghitza, "Auditory Models and Human Performance in Tasks Related to Speech Coding and Speech Recognition", IEEE Transactions on Speech and Audio Processing, 1994, 2(1), 113-131
[32] 焦志平，改进的ZCPA语音特征提取算法研究，[硕士学位论文]，太原，太原理工大学，2005
[33] Siu-Yeung Cho, Tommy W.S.Chow, "Learning Parametric Specular Reflectance Model by Radial Basis Function Network", IEEE Transactions on Neural Networks, 2000, 11 (6), 1498-1503
[34] Martin Hagan, Howard B. Demuth, Mark H. Beale, "Neural Network Design"，北京，机械工业出版社，2002，197-207
[35] 陈逢时，子波变换理论及其在信号处理中的应用，北京，国防工业出版社，1998
[36] G.Ecangelista, S.Cavaliere, "Frequency Warped Filter Banks and Wavelet Transforms: A Discrete-Time Approach Via Laguerre Expansion," submitted to IEEE Trans. on Signal Processing, 1996
[37] G.Ecangelista, S.Cavaliere, "Discrete Frequency Warped WaveletS: Theory and Application," submitted to IEEE Trans.on Signal Processing, special issue on Theory and Application of Filter banks and Wavelets, 1997
[38] G.Ecangelista, S.Cavaliere, "The Discrete-Time Frequency Warped Wavelet Transforms," Proc.of ICASSP'97, Munich, 1997, 2105-2108
[39] Alan V.Oppenheim, Alan S. Willsky(著)，刘树棠(译)，信号与系统，2，陕西，西安交通大学出版社，1998
[40] 王学辉，张明辉，Matlab6.1最新应用详解，北京，中国水利水电出版社，2002
[41] 薛定宇，陈阳泉，高等应用数学问题Matlab求解，北京，清华大学出版社，2004
[42] 赵高峰，基于小波分析的语音端点检测算法研究，[硕士学位论文]，太原，太原理工大学，2006