摘要(英) |
The document combines speech synthesis and speaker conversion and these have relevant application and design. Speech synthesis is that the voice of real man is converted machine voice by synthesis engine. Speaker conversion is based on source speaker and it converts another voice of speaker. To let two techniques can be used in life and entertainment, it needs system to provide implement. The design of the system is that spectrum feather parameter of source speaker is extracted by synthetic software Data of text dependence produced by inputting words and voice model of source speaker input it. And the parameter of target speaker is extracted from voice of target speaker. Both of parameter generate the match table of feather vector of frame by DTW comparing, then GMM is formed by LBG algorithm. After that, using EM algorithm is order to train GMM. When finishing train, parameter correspondence method has transform function. When inputting source spectrum, target spectrum can be got. Besides, synthetic voice of target speaker is formed by speaker is formed by putting pitch feather parameter of source speaker excited and spectrum feather parameter of target speaker together through MLSA (Mel Log Spectrum Approximation). This document proposes many applications and designs of Speech Synthesis and Speaker Conversion. |
參考文獻 |
[1] 蔡松峯, GMM為基礎之語音轉換法的改進,台灣科技大學碩士論文, 2009
[2] Tomoki Toda,Alan W. Black,and Keiichi Tokuda,“Voice Conversion
Based on Maximum-LikelihoodEstimation of Spectral Parameter
Trajectory”,IEEE
[3] Examples for Using Speech Signal Processing Toolkit Ver3.9
http://sourceforge.mirrorservice.org/s/sp/sp-tk/SPTK/SPTK-3.9/SPTK examples-3.9.pdf
[4] REFERENCE MANUAL for Speech Signal Processing Toolkit Ver.3.9
http://sourceforge.mirrorservice.org/s/sp/sp-tk/SPTK/SPTK-3.9/SPTKref-3.9.pdf
[5] 林家緯,使用GMM轉換之背景伴奏消除及趨勢估計之歌曲音高軌跡追蹤 ,交通大學碩士論文, 2013
[6] 白育瑋,使用語言語聲學資訊之高斯混合模型語音轉換應用於可自定
文字轉語音系統, 成功大學碩士論文, 2013
[7] 李昀璋,基於HTS語音合成系統合回歸樹與回朔機制之頻譜係數基頻轉換之語者轉換系統, 成功大學碩士論文,2012
[8] MFCC梅爾倒頻譜係數提取詳解,https://read01.com/JD2NeR.html
[9] 維基百科,梅爾頻率倒譜係數,https://zh.wikipedia.org/wiki
[10] 張智星, Audio Signal Processing and Recognition (音訊處理與辨識),
, 12-2:mfcc
http://mirlab.org/jang/books/audiosignalprocessing/
[11] 張智星,8-4 Dynamic Time Warping
http://mirlab.org/jang/books/dcpr/
[12] GMM:高斯混合模型
http://www.cs.nccu.edu.tw/~whliao/acv2008/08gmm.pdf
[13] GMM:高斯混合模型(Gaussian Mixture Model)
waoffice.ee.kuas.edu.tw /download/建德研究所資料/七月課程/高斯混合模型/高斯混合模型.ppt
[14] 宋柏毅,以韻律模型為基礎之中文韻律轉換研究,交通大學碩士論文,2009
[15] 張智星, Audio Signal Processing and Recognition (音訊處理與辨識),
3-2 Basic Acoustic Features (基本聲學特徵)
http://mirlab.org/jang/books/audiosignalprocessing/
[16] 蔡松峰,使用分段式GMM 及自動GMM 挑選之語音轉換方法, 台灣科技大學碩士論文,2011
[17] 蔡昀庭,基於隱藏式馬可夫模型之中文語音合成系統, 清華大學碩士論文, 2009
[18] 維基百科,梅爾刻度
https://zh.wikipedia.org/wiki/%E6%A2%85%E5%B0%94%E5%88%BB% E5%BA%A6
[19] 賴名彥、蔡松峰,結合HMM頻譜模型與ANN抖音模型之國語歌聲合成,台灣科技大學碩士論文,2014
[20] 露營美工圖案與插圖
https:// shop.xilu.com/product/0002_2013_06_27_13_5780313.html
[21] APP01/大富翁4Fun! 買地炒房現金一把抓 http://fashion.ettoday.net/news/139329
[22] 西遊回合,快樂交友——《神武2》手游
https://read01.com/0kA6jz.html
[23] 啟蒙有聲故事書
https:// shop.xilu.com/product/0002_2013_06_27_13_5780313.html
[24] 玩具咬咬棒(紓壓玩具)
https://tw.bid.yahoo.com/item/100151397123
[25] 張智星, Dynamic Time Warping (DTW), p.7
mirlab.org/jang/books/dcpr/slide/dtw.ppt
[26] Chi-Yueh Lin, National Tsing Hua University, Mandarin TTS using HTS toolkit , p.63
[27] 動態時間校正.ppt, p.3
waoffice.ee.kuas.edu.tw /download/建德研究所資料/七月課程/動態時間校正/動態時間校正.ppt
[28] 父母必讀!這樣說睡前故事,讓孩子一生受益!http://www.spicemami.com/thread-17403-1-1.html
[29] 溫馨童話(雙CD):熊媽媽說故事
http://kingstone.com.tw/book/book_page.asp?kmcode=2015231644518& lid=book_class_sec_se&actid=WISE
[30] 頻域
https://zh.wikipedia.org/wiki/%E9%A0%BB%E5%9F%9F
[31] 蔡佳泓, Interval Estimation, 國立政治大學東亞所, April 15, 2014
http://www3.nccu.edu.tw/~tsaich/EastAsian/Week8_interval2015.pdf |