跳到主要內容

臺灣博碩士論文加值系統

(3.236.124.56) 您好!臺灣時間:2021/07/28 08:26
字體大小: 字級放大   字級縮小   預設字形  
回查詢結果 :::

詳目顯示

: 
twitterline
研究生:劉彥伶
研究生(外文):Yen-Ling Liu
論文名稱:建構華語語音頻率權重分布之分析平台
論文名稱(外文):An Analysis Platform for Frequency Weighting Distribution in Mandarin
指導教授:楊順聰楊順聰引用關係
指導教授(外文):Shuenn-Tsong Young
學位類別:碩士
校院名稱:國立陽明大學
系所名稱:醫學工程研究所
學門:工程學門
學類:生醫工程學類
論文種類:學術論文
論文出版年:2004
畢業學年度:92
語文別:中文
論文頁數:60
中文關鍵詞:頻率權重函數平衡語料語音頻率分布圖
外文關鍵詞:frequency-importancespeech bananaphonetically balanced word
相關次數:
  • 被引用被引用:2
  • 點閱點閱:391
  • 評分評分:
  • 下載下載:58
  • 收藏至我的研究室書目清單書目收藏:0
對人類來說,語音頻率特性對語音的辨識占有很重要的地位,學者早就對英文找出其完整之語音頻率分布,並有許多人在研究頻率重要性函數,試圖找出何種頻帶對於人們辨識較為重要。語音頻率權重分布圖可做為助聽器調適以及人工電子耳編碼策略選擇的依據,而根據英文的語音分布來做助聽器的補償,並不真正的適合我們,因此我們試圖建構一套平台,找出屬於華語的語音能量分布,並加上我們定義的頻率權重函數,來預測華語音素在各頻帶之重要性。此平台主要包括四個部分,包括錄音模組、頻率分析模組、權重統計模組,以及頻率權重顯示模組。針對語音頻率權重分析材料來源部分,我們利用注音符號的聲母及韻母共59 個,作為華語語音的代表,找出華語語音的頻率分布情形;在分析模組部分,主要是使用1/3 Octave 的分頻方式,在人耳所能聽到的範圍切出30 個頻帶,以求出每個語音在各個頻帶的能量分布,再由中研院漢語平衡語料庫3.0所統計各音素出現之機率,來做為每個語音之權重依據,將之與每個頻帶能量做權重調整,如此便可得到每個語音頻率權重分布。若將每個音素對應語音所導致的語音頻率權重分布做統計,便可得到最後的華語語音頻率權重分布。透過此一平台不止可以了解到華語各音素的分布情形,及整個華語的分布情形,也能藉由此平台去預估何種頻帶對於我們辨識較為重要,更能夠做為助聽器調適以及人工電子耳編碼策略選擇的依據,讓聽力損失的人得到最好的幫助。
Many factors can affect speech recognition, but the frequency distribution of speech is considered as the most important one. The frequency distribution of speech in English has been proposed by some researchers years ago. Many literatures focus on frequency-importance functions, and these people try to find which band is important for people to recognize. The strategy selection for hearing aid fitting and cochlear implant coding heavily depends on the frequency distribution of speech.Most available frequency distribution of speech, derived from English, is not really
compatible with the application for Mandarin speakers. We developed a platform to analyze the frequency distribution of spoken Mandarin and propose our frequency weighting function. The platform comprised of four functional modules; a waveform-recording module, a speech analysis module, a statistics module and a plot weighting distribution module. We used 59 basic phonemes of Mandarin as the material to analyze the frequency distribution. A filter bank with 1/3 octave central frequency was built to split the energy of the phonemes into 30 different frequency
bands. The Sinica Corpus 3.0 was used to get weighting of the phonemes. By multiplying together the energy of each frequency band and the phoneme weighting,we can get the frequency weighting distribution of each phoneme. When gathering the frequency weighting distributions of all phonemes, we can plot out the whole frequency weighting distribution of Mandarin and propose it for the application in hearing rehabilitation.
[1] 林寶貴“語言障礙與矯治"五南書局,p.19-21,民91
[2] 葛本儀“語言學概論" 五南書局,p.118-128,民91
[3] Winckel, F., “Acoustical foundations of phonetics” Manual of Phonetics B.Malmberg(ed), NorthHolland, p.17-44, 1968
[4] http://www.oticon.com.tw/human/human1.htm
[5] Valente, M., Maureen V., Handbook of Clinical Audiology: chapter 37 Hearing Aid Fitting and Verification Procedures for Adults., p.714, 2002.
[6] Ay, S. U., Zeng, F. G.., Sheu, B. J., “Hearing with Bionic ears-speech processing strategies for cochlear implant devices” IEEE Circuits and Devices Magazine,May, p.18-23, 1997.
[7] Loizou, P., "Signal-processing techniques for cochlear implants" IEEE Engineering in Medicine and Biology Magazine,18(3), p.34-46, 1999.
[8] DePaolis, R. A., Janota, C. P., Frank, T., ” Frequency importance functions for words, Sentences, and Continuous Discourse” Journal of Speech and Hearing Research, 39, p.714-723, 1996.
[9] Studebaker, G. A., Sherbecoe, R. L., “Frequency-importance functions for speech recognition” Chapter in Studebaker, G. A. and Hochberg, I. (Eds.) Acoustical
Factors Affecting Hearing Aid Performance (Second Edition). Allyn and Bacon,Needham Heights, MA., 1993.
[10] 王老得、蘇富美“中國語音均衡字彙表之編製研究"耳鼻喉科醫學會雜誌,14(2),p.87-95,1979
[11] 余秀敏、劉繼謚“國語語音特性平衡句之建立"電信研究季刊,19(1),p.79-85,民78
[12] Studebaker, G. A., Pavlovic, C. V., Sherbecoe, R. L., ”A frequency importance function for continuous discourse” Journal of the Acoustical Society of America,
81, p.1130-1138, 1987.
[13] Duggirala, V., Studebaker, G. A., Pavlovic, C. V., Sherbecoe, R. L., ” Frequency importance functions for a feature recognition test material“ Journal of the
Acoustical Society of America, 83, p.2372-2382, 1988.
[14] Steeneken, H. J. M., Houtgast, T., ”Mutual dependence of the octave-band weights in predicting speech intelligibility” Speech Communication, 28(2),p.109-123, 1999.
[15] Steeneken, H. J. M., Houtgast, T., ”Phoneme-group specific octave-band weights in predicting speech intelligibility” Speech Communication, 38(3-4), p.399-411,
2002.
[16] Sherbecoe, R. L., Studebaker, G.. A., “ Audibility-index functions for the connected speech test"Ear Hear, Oct, 23(5), p.385-98, 2002.
[17] Mueller, H. G., Killion, M. C., ”An Easy Method For Calculating the Articulation index” The Hearing Journal, 43(9), p.14-17, 1990.
[18] Chris, H., Aaron, T., Zezhang, H., ”The articulation index in clinical diagnosis and hearing aid fitting” Current Opinion in Otolaryngology & Head and Neck Surgery, 4, p.325-334, 1996.
[19] Pavlovic, C.V., ”Band Importance Functions for Audiological Application” Ear
and Hearing, 15(1), p.100-104, 1994.
[20] 葉德明"國語注音符號教材"國立臺灣師範大學國語教學中心,1993
[21] 張錦松,韓光榮,張錦輝編著” 噪音振動控制 ”p.42-44;p.66,高立圖書,
民92
QRCODE
 
 
 
 
 
                                                                                                                                                                                                                                                                                                                                                                                                               
第一頁 上一頁 下一頁 最後一頁 top