臺灣博碩士論文加值系統

支氣管呼吸音是由於口鼻吸入或呼出的氣流在聲門及氣管或主支氣管所形成湍流的聲音。它不僅可做為呼吸道或肺部疾病的判斷，也因隱藏了個人獨特生理構造的資訊，故也應可識別出個人的身分，從而區別不同的病人。本研究透過聽診器來收錄頸部所呈現之支氣管呼吸音，然後計算其梅爾頻率倒頻譜係數，並將每個人之呼吸音的梅爾頻率倒頻譜係數表示成高斯混合模型，做為未知身分之呼吸音的判斷依據。我們也進一步利用「i-vector+LDA」方法提升身分識別的正確率，在我們所收集之8個測試者的資料庫中進行實驗，結果可達90%的識別正確率。

Bronchial breath sound is the sound of turmoil flow produced by the inspirated air through glottis, trachea or major bronchi. It can not only be used to determine the respiratory tract and lung related diseases, but also to distinguish one person from the other and thereby identifying patients, since it contains personal physiological characteristics. This study captures the bronchial breath sound using a stethoscope attached on a subjects neck. For each person, the mel frequency cepstral coefficients (MFCCs) are computed for the captured sounds, and then further represented by a Gaussian mixture model (GMM). Given an unknown breath sound recording, we can determine who among a set of candidate people produced the breath sound by matching the MFCCs of the sound to each of the GMMs. Furthermore, we apply i-vector approach to boost the identification accuracy. Our experiments conducted on a dataset composed of 8 persons show that the accuracy of identifying people from their breath sounds can be 90%.

摘要 i
ABSTRACT ii
致謝 iii
目錄 iv
表目錄 viii
圖目錄 ix
第一章 緒論 1
1.1 研究動機 1
1.2 呼吸音介紹 2
1.2.1 肺泡呼吸音 2
1.2.2 支氣管呼吸音 3
1.2.3 呼吸音比較 3
1.3 研究方法 4
1.4 論文結構 5
第二章 實驗工具及量測位置 6
2.1 實驗工具 6
2.2量測位置 7
2.3呼吸音量測結果 8
第三章 高斯混合模型的呼吸音辨識 9
3.1特徵參數擷取 10
3.1.1 預強調 11
3.1.2 音框化 11
3.1.3 漢明窗 12
3.1.4 快速傅立葉轉換 12
3.1.5 梅爾刻度三角帶通濾波器 13
3.1.6 對數轉換 14
3.1.7 離散餘弦轉換 14
3.2高斯混合模型 15
3.2.1 呼吸音模型訓練流程 16
3.3身份識別 18
第四章 基於LDA+i-vector模型呼吸音辨識 19
4.1前處理 21
4.1.1 平均值消去 22
4.1.2 變異數正規化 23
4.1.3 ARMA濾波器 23
4.2語者模型建立 24
4.2.1 i-Vector模型 24
4.2.2 波式統計量 25
4.2.3 LDA分類器 26
4.3驗證分數計算 29
第五章 實驗結果與討論 30
5.1情境一 不同換氣方式的辨識比較 31
5.1.1 鼻吸鼻吐 32
5.1.2 嘴吸嘴吐 34
5.1.3 鼻吸嘴吐 36
5.1.4 嘴吸鼻吐 38
5.1.5 情境一 實驗整理與討論 40
5.2情境二 呼吸模式的辨識比較 41
5.2.1 自然呼吸與提示呼吸能量分析 42
5.2.2 GMM實驗 44
5.2.3 LDA實驗 45
5.2.4 情境二 實驗整理與討論 46
5.3情境三 脈搏音對呼吸音身份識別之影響 47
5.3.1 脈搏音身分識別 48
5.3.2 純呼吸之身分識別 49
【GMM實驗】 49
【LDA實驗】 51
5.3.3 情境三 實驗整理與討論 53
第六章 結論與未來展望 54
6.1 結論 54
6.2 未來展望 55
參考文獻 56

[1]. Reza Darooei, Ali Mahloojifar, “Asthma Diagnosis Based on Respiratory Dynamic Using Sparse Representation Based-Classifier” IEEE, 22nd Iranian Conference on Biomedical Engineering (ICBME),pp. 216-220, 2015
[2]. Eugenijus Kaniusas, Helmut Pfützner, and Bernd Saletu,“Acoustical Signal Properties for Cardiac/Respiratory Activity and Apneas”IEEE Transactions on Biomedical Engineering, Vol:52,pp. 1812-1822, 2005
[3]. Ria Lestari Moedomo, M. Sukrisno Mardiyanto, Munawar Ahmad,“The Breath Sound Analysis For Diseases Diagnosis and Stress Measurement”IEEE System Engineering and Technology (ICSET),pp. 1-6, 2012
[4]. Yanzhi Ren, Chen Wang, Jie Yang,“Fine-grained Sleep Monitoring: Hearing Your Breathing with Smartphones”IEEE Conference on Computer Communications(INFOCOM),pp. 1194-1202, 2015
[5]. Phil Corbishley and Esther Rodríguez-Villegas,“Breathing Detection: Towards a Miniaturized,Wearable, Battery-Operated Monitoring System” IEEE TRANSACTIONS ON BIOMEDICAL ENGINEERING, VOL. 55, 2008
[6]. Dehak, N., Kenny, P., Dehak, R., Dumouchel P., and Ouellet P., “Front-End Factor Analysis for Speaker Verification,” IEEE Transactions on Audio, Speech and
Language Processing, vol. 18, no. 8, pp. 1-28. Jul. 2010.
[7]. P. Matejka, O. Glembek, F. Castaldo, O. Plchot, P. Kenny, L. Burget, and J. Cernocky, “Full-covariance ubm and heavy-tailed plda in i-vector speaker verification,” Proc. ICASSP ’11, pp. 4828–4831, 2011.
[8]. J.M.K. Kua, J. Epps, E. Ambikairajah,“i-vector with sparse representation classification for speaker verification, ” Speech Commun, 2013.
[9]. Chia-Ping Chen and Jeff A. Bilmes,“MVA Processing of Speech Features” IEEE Transactions on Audio, Speech, and Language Processing ,Vol:15,pp. 257-270, 2007
[10]. A. Hatch, S. Kajarekar, and A. Stolcke, “Within-Class Covariance Normalization for SVM-Based Speaker Recognition,”in International Conference on Spoken Language Processing, Pittsburgh, PA, USA, pp. 1471-1474,September 2006.
[11]. Snack Sound Toolkit, http://www.speech.kth.se/snack/
[12]. Xing Fan and John H. L. Hansen,“Speaker Identification Within Whispered
Speech Audio Streams”IEEE TRANSACTIONS ON AUDIO, SPEECH, AND LANGUAGE PROCESSING,VOL.19,NO5,2011