研究成果

本論文已實作完成一以USB1.1 為傳輸介面之 8 通道及時性麥克風陣 列訊號處理平台，並在此平台實作完成真人語音偵測（VAD）與適應性空 間濾波器整合，可使用於室內環境或車內環境，測試結論為：

1. 於吵雜環境下使用麥克風陣列，其語音辨識率較使用單一麥克風高。

2. 空間濾波器階數的提高可提升其SNR。

3. VAD 與空間濾波器的結合可應用於噪音源位置變動的環境，若只使用 空間濾波器則噪音源位置改變，則須人為去重新啟動LMS，使其重新 調整空間濾波器係數，若加入VAD 則會當沒真人語音時，系統自動調 整空間濾波器係數。

VAD 與空間濾波器的結合其語音辨識率高於只使用空間濾波器。

5. VA

6. 在車內環境中，引擎一啟動就有一低頻雜訊，可透過空間濾波器將此 低頻雜訊濾除乾淨，但若麥克風增益大時，空間濾波器無法消除路面 顛簸的影響，但可透過VAD 消去路面點跛的影響。

.2 未來展望

在做LMS 演算法時，必須先針對雜訊作出 zero-mean 及 Gaussian 的假

，若雜訊本身並非zero-mean 及 Gaussian，則 Normalize LMS 處理效果 有限，但現實生活中並非所有接收到的雜訊皆為zero-mean 及 Gaussian，

因此未來日子可考慮其他適應性訊號處理演算法，像H∞等。

4.

D 結合空間濾波器適用於噪音源變動的環境。

6

設 會

Reference

[1] D. Johnson and D. Dudgeon, Array Signal Processing: Concepts and [2] J. L. Flanagan﹐ L. Landgraf﹐ D. J. McLean﹐ ＂Matched-filter

[3] Barry D. Van Veen and Kevin M.Buckley,”Beamforming:A Versatile rray signal processing in the beamspace domain: novel techniques based on the poor resolution of

[5] Byung-Chul Kim; I-Tai Lu , ”High resolution broadband beamforming e: 3 , pp. 1673 –1676, 2000

iano, J.S., Jr.; Vu, T.B., ”Reduced complexity MVDR broadband equency invariant constraint,” Antennas

00. IEEE , Volume:

[7] Pillai, S. Unnikrishna, Array signal processing, 1989 [8]

Antennas and Propagation, IEEE Transactions on , Volume: 46 Issue: 5 , pp. 609 –617, May 1998

; Yih-Fang Huang , “Optimally combined nonlinear MMSE beamforming and interference cancellation for CDMA communications,”

[10] 黃佑霖,”應用延遲濾波器之麥克風陣列訊號處理”,交大電控碩士論 文,Jun 2000

[12] ,”應用於個人電腦環境之即時語音純化系統設計”,交大電控碩 士論文,Jul 2004

[13] Ta-Sung Lee,Array Signal Processing,(class note)

Techniques,Prentice Hall, Englewood Cliffs, New Jersey, 1993.

processing of hydrophone array ＂ ﹐ J. Acousr. Soc. Am. 42, 1165 (A)(1967).

Approach to Spatial Filtering,”IEEE ASSP MAGAZINE April 1988.

[4] Zoltowski, M., ”High resolution sensor a

Fourier beamforming,” Spectrum Estimation and Modeling, 1988., Fourth Annual ASSP Workshop on , pp. 350 –355, 1988

based on the MVDR method,” OCEANS 2000 MTS/IEEE Conference and Exhibition , Volum

[6] Marc

beamspace beamforming under fr

and Propagation Society International Symposium, 20 2 , pp. 902 –905, 2000

Ta-Sung Lee; Tsui-Tsai Lin , “Coherent interference suppression with complementally transformed adaptive beamformer,”

[9] Gollamudi, S.

Personal Wireless Communications, 2000 IEEE International Conference on , pp. 474 –478, 2000

文,Jun 1999

[11] 陳界全,”即時聲源追蹤與空間濾波器設計”,交大電控碩士論 康創閔

[14] Javier Ramírez , José C. Benítez , Ángel de la Torre

and Antonio Rubio ,”E ity detection algorithms using long-term speech information,” Speech Communication, Volume 42,

[16] European fficient voice activ Issues 3-4, April 2004, Pages271-287

[15] European Digital Cellular Telecommuni- cations System; Half rate speech; Voice Activity Detection (VAD), ETSI GSM 06.42 (ETS 300-581-6), 1995.

Digital Cellular Telecommuni- cations System; Half rate speech; Half rate speech transcoding, ETSI GSM 06.20 (E 300-581-2), 1995.

[17] ITU-T G.729, Coding of Speech at 8kbit/s Using CS-ACELP, March, 1996.

A. Benyassine, E. Shlomot, H. Y. Su, D. Massaloux, C. Lamblin, and J. P. Petit, “ITU recommendation G.729 annex B: A silence compression scheme for use with G.729 optimized for V.70 digita simultaneous voice and data applications,” IEEE Commun. Mag., vo 35, pp. 64–73, Sept. 1997.

Ali H. Sayed, Fundamentals of Adaptive Filtering,

[20] Dahl, M.; Claesson, I., “Acoustic noise and echo cancelling with microphone array,” Vehicular Technology, IEEE Transactions on , Volume: 48 Issue: 5 , Sept.1999 Page(s): 1518 –1526

來源網站：http://andr

[22] Xuedong Huang,Alex Acero and Hsiao-Wuen Hon,Spoken Language [23]

Processing.

來源網站： http://www-306.ibm.com/software/voice/viavoice/

[24]

[25]

[26]

Cypress Semiconductor, The EZ-USB FX Technical Reference Manual, Version 1.2.Cypress Semiconductor Corporation, 2000.

許永和,8051 微處理機程式設計,pp3-11 – 3-15.

來源網站： http://www.intel.com/software/products/mkl/

[27] Bradley Bargen and Peter Donnelly, Inside DirectX, 1999, pp. 203-246.

[28] 來源網站： http://speech.cm.nctu.edu.tw/