臺灣博碩士論文加值系統

語音傳輸在多媒體電信服務上是最為常見的，特別是透過網際網路進行語音通訊更是現今一個新興的技術，但高位元率的語音串流在有限的網路頻寬資源，將產生通話斷訊導致語音品質下降。為了讓語音通話滿足“連續性”的基本需求，必須將語音訊號使用高壓縮率編碼以產生低位元率的資料串流，如此才能在擁塞的網路環境下維持連續通話，而採用低位元率編碼器必需要兼顧到語音品質和計算複雜度。國際電信聯盟(ITU-T)所訂定的語音編碼標準G.723.1與G.729被認為是最好的低位元率語音標準，並同屬於CELP之合成再分析編碼結構，雖具有良好的語音品質和低位元率之優點，但其令人詬病的是需要較高的計算量，因此有必要對編碼演算法進行研究與改善，以達到降低計算複雜度且保持語音品質。
語音編碼效率之改善，本論文對於G.723.1在週期性激發源部分之適應性基週增益向量碼簿搜尋演算法，利用三階開迴路殘值訊號與一階閉迴路預估器來預測最佳的五階閉迴路基週增益向量。另外在非週期性之激發脈衝及增益編碼於G.723.1編碼器是採用MP-MLQ/ACELP演算法搜尋隨機碼簿之碼向量，MP-MLQ部分，則參考ACELP激發碼簿結構，分別對奇、偶子音框設計適合的碼簿結構，並以脈衝能量函數對每軌進行篩選較佳脈衝位置。ACELP部分，以脈衝能量函數對每軌篩選較佳之脈衝位置，再以深度優先樹搜尋，並運用於G.729之隨機碼簿搜尋。由於兩編碼器同屬於CELP編碼架構，因此在相互通訊時對於線頻譜對和基週之參數，本論文提出以線性內插技術處理。在兩編碼器相互通訊時，可進一步減少編碼過程的計算複雜度，本文所提出的快速演算法同時實現於G.723.1與G.729編碼器均可保持一定的語音品質，人耳無法察覺品質衰減。

Speech communication is the most common service in the Internet telecommunication and multimedia process. However, since speech signal should be continuously sent back, the voice in the service of the Internet should collect enough speech data, which can cause large speech delay and can degrade the speech quality in a limited network bandwidth. To achieve "continuity", speech codec with high compression rate has been used to generate a low-rate data stream, but that codec requires higher computational complexity. Thus, reducing the bit rate and improving speech quality of codec is the most significant. ITU-T offers the G.723.1and G.729 codecs that have used popularly in the Internet applications. These codecs offer high quality and low bit rate coding constitution.
This paper predict the search range of adaptive codebook-gain in the G.723.1 standard codec by minimizing the mean square error between the three-tap excitation signal with its residual signal and one-tap pitch predictor. For the G.723.1 MP-MLQ, we propose a fast search algorithm by using a designed energy function and the multi-track positions structure of the stochastic excitation signals to predict the candidate pulses for each subframe. As for both of the G.723.1 and the G.729 ACELP codebook, we base on depth-first tree search (DFS) and pulse-position likelihood-estimate to propose a fast search algorithm. As the two encoders belong to CELP coding structure, transcoding procedures are completed through two processes: line spectral pair and pitch conversions. They are all used to linear interpolation processing. For further computational complexity reduction, we use two fast search algorithms. First, we employ residual signals to predict candidate gain-vectors of adaptive-codebook in the G.723.1. Next, we adopt fast stochastic excitation pulses search method. Simulation results show that the proposed methods reduce a large amount of computation. Also, reconstructed speech signal still maintain a certain level of speech quality with perceptually negligible degradation.

摘要 i
英文摘要 ii
致謝 iii
目次 iv
表目錄 vii
圖目錄 ix
第一章 緒論 1
1.1 語音壓縮介紹 1
1.2 語音壓縮標準之演進 2
1.3 論文研究之目的 3
1.4 論文大綱 4
第二章 G.723.1與G.729語音編碼器 5
2.1 ITU-T G.723.1語音編碼器 6
2.1.1 前處理 7
2.1.2 線性預估分析(LPC Analysis) 8
2.1.3 共振峰聽覺感官加權濾波器(FPWF) 10
2.1.4 諧波雜訊濾波器(Harmonic Noise Shaping) 10
2.1.5 脈衝響應計算(Impulse Response Calculator) 11
2.1.6 零輸入響應(Zero Input Response) 12
2.1.7 開迴路基週搜尋(Pitch Estimator) 12
2.1.8 閉迴路基週搜尋(Pitch Predictor) 13
2.1.9 代數碼激式線性預測(ACELP) 15
2.1.10 多脈衝最大相似度量化搜尋(MP-MLQ) 17
2.1.11 G.723.1語音編碼器位元分配 18
2.2 G.723.1與G.729語音編碼器之差異 19
2.3 語音品質測量方法 22
2.3.1 主觀評估法(Subjective Quality Measure) 22
2.3.2 客觀評估法(Objective Quality Measure) 22
第三章 有效率的快速編碼演算法 24
3.1 G.723.1適應性碼簿之快速搜尋演算法 25
3.1.1 三階開迴路預測五階閉迴路基週增益 26
3.1.2 五階閉迴路基週增益快速預估演算法 29
3.2 G.723.1隨機碼簿之快速搜尋演算法 35
3.2.1 脈衝能量函數預選法 36
3.2.2 雙多軌結構預選法基於類ACELP 38
3.2.3 深度優先樹搜尋 42
3.2.4 簡化的ACELP架構採用深度優先搜尋編碼演算法 43
第四章 G.723.1與G.729相互通訊研究 48
4.1 G.729至G.723.1通訊 49
4.1.1 G.729至G.723.1之線頻譜對處理 50
4.1.2 G.729至G.723.1之基週處理 51
4.2 G.723.1至G.729通訊 53
4.2.1 G.723.1至G.729之線頻譜對處理 54
4.2.2 G.723.1至G.729之基週處理 55
第五章 實驗結果及效能之評估 58
5.1 快速搜尋法之實驗 58
5.1.1 三階開迴路預測五階閉迴路基週增益實驗 58
5.1.2 脈衝能量函數預選法實驗 61
5.1.3 簡化的ACELP架構採用深度優先搜尋編碼演算法實驗 62
5.2 G.723.1與G.729相互通訊之實驗 64
5.2.1 客觀評估 64
5.2.2 主觀評估 65
5.2.3 計算複雜度評估 65
5.3 整體效率評估 67
第六章 結論 68
參考文獻 69
附錄 72
A. G.723.1語音編碼器之殘值預估激發訊號快速搜尋演算法 72
B. Low Complexity Multi-track Search Scheme for G.723.1 MP-MLQ 78
C. ACELP語音編碼器激發脈衝位置快速搜尋演算法之分析與研究 82
作者簡介 88

[1]王小川，語音訊號處理，全華科技圖書股份有限公司，2005。
[2]李昌立、吳善培，數字語音—語音編碼實用教程，人民郵電出版社，2004。
[3]ITU-T Rec. G.723.1, Dual Rate Speech Coder for Multimedia Communications at 5.3 and 6.3 kbit/s, March 1996.
[4]ITU-T Rec.G.729, Coding of Speech at 8 kbit/s using Conjugate Structure Algebraic Code Excited Linear Prediction (CS-ACELP), 1996
[5]ITU-T Rec. H.323, Visual Telephone Systems and Equipment for Local Area Networks Which Provide a Non guaranteed Quality of Service, March 1996.
[6]林裕斌, “ITU-T G.729 和G.723.1 語音編碼器之快速演算法,” 國立成功大學電機工程研究所碩士論文, 2002。
[7]S. Ruslan, “Method of Converting Speech Codec Formats between G.723.1 and G.729А,” CADSM’2007, pp. 20-24, February 2007.
[8]ITU-T Rec. P.800, Methods for Subjective determination of Transmission Quality, August 1996.
[9]ITU-T Rec. P.862, Perceptual Evaluation of Speech Quality (PESQ), an Objective Method for End-to-end Speech Quality Assessment of Narrowband Telephone Networks and Speech Codecs, February 2001.
[10]S. M. Lee, S. Park and Y. Jang, “Cost-effective Implementation of ITU-T G.723.1 on A DSP Chip,” Proceedings of 1997 IEEE International Symposium on Consumer Electronics, pp. 31-34, December 1997.
[11]F. K. Chen, and J. F. Yang, “Candidate Scheme for MP-MLQ Search in G.723.1,” proceeding of third IEEE Signal Processing Workshop on Signal processing advances in wireless communications, Taoyuan, Taiwan, pp. 386-371, March 2001.
[12]S. K. Jung, Y. C Park, S.W. Youn, K.T. Kim, and D.H. Youn, “Efficient Implementation of ITU-T G.723.1 Speech Coder for Multichannel Voice Transmission and Storage,” Eurospeech 2001 – Scandinavia, 2001
[13]S. K. Jung, K. T. Kim, Y. C. Park, and H. G. Kang, “A Fast Adaptive-Codebook Search Algorithm for G.723.1 Speech Coder,” IEEE Signal processing letters, vol. 12, no.1, pp. 75-78, January 2005.
[14]陳育昌、林榮三、陳福坤, “ITU-T G.723.1 語音編碼器適應性增益碼簿不均分段之快速搜尋演算法,” 2007資通技術管理與應用會議, 六月 2007。
[15]林榮三、陳穎澈、陳福坤, ”殘值訊號預估 G.723.1適應性增益碼簿快速搜尋演算法” , 全國計算機研討會, 2007。
[16]C. Zhang, K. Tang, H. Cui, W. Du, and J. Li, “Efficient pitch predictor algorithm in ITU-T G.723.1,” IEEE Conf. Communication Technology Proceedings, Vol. 02, pp. 1727-1729, April 2003.
[17]C. Negrescu, ” Optimization algorithm for the MP-MLQ excitation in G723.1 encoder," IEEE Conf. Electronics, Circuits and Systems, pp. 1003-1006, June 2006.
[18]林榮三、陳穎澈、陳福坤, “G.723.1 語音編碼器之殘值預估 MP-MLQ快速搜尋演算法,” 2008數位生活科技研討會-人本智慧生活科技, 六月 2008。
[19]R. S. Lin, Y. C. Chen, and F. K. Chen, “Lower Complexity Search Method for G.723.1 MP-MLQ Algorithm,” The eighth International Conference on Intelligent Systems Design and Applications, pp.158-161, November 2008.
[20]Y. C. Chen, R. S. Lin, F. K. Chen, “Complexity Reduction of the Stochastic Code- Vector Search for ITU-T G.723.1 Codec,” Workshop on Consumer Electronics, pp. 44, November 2006.
[21]ITU-T Recommendation G.729.1, An 8-32 kbit/s Scalable Wideband Coder Bitstream Interoperable with G.729, May 2006.
[22]ITU-T Rec. G.729-Annex A, Reduced Complexity 8 kbit/s CS-ACELP Speech Codec, 1996.
[23]S. F. Campos Neto and F. L. Crcoran, “Performance Assessment of Tandem Connection of Enhanced Cellular Coders,” IEEE Proceedings of International Conference on Acoustics Speech Signal Processing, pp. 177–180, March 1999.
[24]H. G. Kang, H. K. Kim, H. Cui, W. Du and R. V. Cox, “Improving transcoding capability of speech coders in clean and frame erasured channel environments,” In: Proceedings of IEEE Workshop on Speech Coding, pp. 78-80, September 2000.
[25]S. W. Yoon, S. K. Jung, Y. C. Park, and D. H. Youn, “An Efficient Transcoding Algorithm for G.723.1 and G.729A Speech Coders,” Proceedings of Eurospeech, pp. 2499–2502, September 2001.
[26]K. T. Kim, S. K. Jung, Y. C. Park, Y. S. Choi and D. H. Youn, “An Efficient Transcoding Algorithm for G.723.1 and EVRC Speech Coders,” IEEE Conf. VTS 54th, vol. 3, pp. 1561–1564, October 7-11, 2001.
[27]S. Seo, D. Jang, S. Lee and Chang D. Yoo, “A Novel Transcoding Algorithm for SMV and G.723.1 Speech Coders via Direct Parameter Transformation,” Eurospeech 2003 – CENEVA, pp. 2861–2864, September 1-4, 2003.
[28]S. Lee, S. Seo, D. Jang and Chang D. Yoo, “A Novel Transcoding Algorithm for AMR and EVRC Speech Coders via Direct Parameter Transformation,” IEEE Conf. ICASSP’03, vol. 2, pp. II–177–80, April 6-10, 2003.
[29]S. W. Yoon, H. G. Kang, Y. C. Park, and D. H. Young, “An Efficient Transcoding Algorithm for G.723.1 and G.729A Speech Coders: Interoperability between Mobile and IP Network,” Speech Communication, vol. 43, no 1-2, pp. 17-31, June 2004.
[30]M. Ghenania and C. Lamblin, “Low-Cost Smart Transcoding Algorithm between ITU-T G.729 (8 kbit/s) and 3GPP NB-AMR (12.2 kbit/s),” Eurospeech 2004, pp. 1681–1684, September 6-10, 2004.
[31]J. K. Choi, C. H. Lee, H. G. Kang, Y. C. Pang and D. H. Youn, “Improvement Issues on Transcoding Algorithms : for The Dlexible Usage to The Various Pairs of Speech Codec,” IEEE Conf. ICASSP’04, vol. 1, pp. I–269–72, May 2004.
[32]C. Lamblin and M. Ghenania, “An Efficient Method for Multi-Tap Long Term Predicitor (LTP) Transcoding : Application to ITU-T G.723.1,” Eusipco 2006, September 4-8, 2006.
[33]S. Ruslan, H. Ludmila and B. Pavlo, “Method of Converting Speech Codec Formats between GSM 06.20 and G.729,” IEEE International Workshop on Intelligent Data Acquisition and Advanced Computing Systems: Technology and Applications 6-8, pp. 686-689, September 2007.
[34]C. Beaugent, “Smart Transcoding between CELP Speech Codecs Through Voiced Oriented Pitch Mapping,” Proceedings of IEEE Workshop on MMSP, pp. 155-158, October 1-3, 2007.
[35]G. P. Kwon, J. M. Chung and S. J. Nam, “A Novel Transcoding Technique between EVRC and G.729A for Mobile Multimedia Devices,” IEEE Transactions on Consumer Electronics, vol. 53, Issue 3, pp. 885-890, August 2007.
[36]S. K. Wang, Z. H. Yang and Z. Y. Wu, “Transcoding Scheme between AMR-WB and VMR-WB”, IEEE Conf. CISP’04, pp. 1-5, October 17-19, 2009.
[37]Z. Y. Liu, C. C. Bao, J. W. Qiu and H. Xu, “LSP Transcoding Method between 3GPP AMR-NB and ITU-T G.729A Based on GMM,” Journal on Communications, vol. 31, no. 2, pp. 44–50, February 2010.