臺灣博碩士論文加值系統

本論文提出一個在缺乏中文-英文混合語音語料庫的狀況下建立中文-英文混合語音合成器的方法。由於研究資源的限制，在建立TTS時有兩個由同一位語者錄製的中文和英文語料庫可以使用，建立出兩個系統：一個簡易的基本系統和一個進階的系統。基本系統由基於HMM的語音合成系統(HTS)直接建構，使用兩個獨立語音資料庫以及合併的語言參數來訓練。進階系統則是通過考慮了中文與英文分享同樣的斷點資訊以及聲調資訊來訓練。兩種系統通過主觀的聽覺測試來比較，為應用於語音合成系統上的未來研究打下基礎。

This paper presents the construction of a Mandarin-English mixed text-to-speech system (ME-TTS) without a Mandarin-English mixed speech corpus. Due to the limitation of the research resources, only two independent Mandarin and English corpora uttered by the same speaker are available for the construction of the ME-TTS. Two ME-TTSs are constructed: a naïve baseline system and an advanced system. The naïve baseline system is directly constructed by the HMM-based speech synthesis system (HTS) with the two speech corpora and their language-dependent linguistic features. The advanced system is constructed by considering a shared tone and break system among Mandarin and English. The two ME-TTSs are evaluated by subjective tests to show the performances and pave the path for the future studies.

第一章 緒論
第二章 簡易基本系統
第三章 進階系統
第四章 系統評估
第五章 結論
參考文獻
附錄

1.T. Yoshimura, K. Tokuda, T. Masuko, T. Kobayashi, T. Kitamura, "Simultaneous Modeling of Spectrum Pitch and Duration in HMM-Based Speech Synthesis", Proc. of EUROSPEECH, vol. 5, pp. 2347-2350, 1999.
2.K. Tokuda, T. Yoshimura, T. Masuko, T. Kobayashi, T. Kitamura, "Speech parameter generation algorithms for HMM-based speech synthesis", Proc. of ICASSP 2000, vol. 3, pp. 1315-1318, June 2000.
3.K. Tokuda, T. Kobayashi, T. Masuko, and S. Imai, “Mel-generalized cepstral analysis – A unified approach to speech spectral estimation”, Proc. ICASSP, pp.1043-1046, 1994.
4.T. Kobayashi and S. Imai, “Spectral analysis using generalized cepstrum,” IEEE Trans. Acoust., Speech, Signal processing, vol. ASSP-32, pp.1087-1089, Oct. 1984.
5.David Talkin “A Robust Algorithm for Pitch Tracking (RAPT)” in Speech Coding and Synthesis, W. B. Kleijn and K. K. Palatal (eds), pages 497-518, Elsevier Science B.V., 1995
6.The Speech Signal Processing Toolkit (SPTK) SPTK-3.11 source code and demonstrations. [Online] . Available: http://sp-tk.sourceforge.net/
7.HTS Working Group. (2017, Jan.). HTS-2.3 source code and demonstrations. [Online]. Available: http://hts.sp.nitech.ac.jp/?Download
8.John J. Dubnowski, Ronald W. Schafer , and Lawrence R. Rabiner, "Real time digital hardware pitch detector," IEEE Trans. Acoust., Speech, Signal Process, vol. 24, no. 1, pp. 2-8, Feb 1976
9.S. Young , G. Evermann , M. Gales , T. Hain , D. Kershaw , G. Moore , J. Odell , D. Ollason , D. Povey , V. Valtchev , P. Woodland , “The HTK Book”, Cambridge University , 2005
10.Y.-R. Wang, The NCTU Speech Lab Traditional Chinese Parser. NCTU Speech Lab, 2017. Available: http://parser.speech.cm.nctu.edu.tw/.
11.Speech at CMU. (2017, Jan.). The CMU Pronouncing Dictionary. [Online].http://www.speech.cs.cmu.edu/cgi-bin/cmudict
12.L.-P. Chang and K.-J. Chen, “The CKIP Part-of-speech Tagging System for Modern Chinese Texts,” in Proc. of ICCPOL 1995, Hawaii,1995.
13.Mitchell P. Marcus, Beatrice Santorini, and Mary A. Marcinkiewicz, “Building a large annotated corpus of English: The Penn Treebank.” Computational Linguistics, vol. 19 , pp. 313-30, 1994.
14.Available at http://www.aclclp.org.tw/use_mat_c.php#ntut
15.Yih-Ru Wang , Sin-Horng Chen ,“Tone Recognition of Continuous Mandarin Speech Based on Neural Networks,”IEEE Trans. On Speech and Audio Processing ,Vol.3,No2,pp.146-150.March 1995.
16.C.-Y. Chiang, S.-H. Chen, H.-M. Yu, and Y.-R. Wang, “Unsupervised joint prosody labeling and modeling for Mandarin speech,” J. Acoustic Society of America., vol. 125, no. 2, pp. 1164-1183, Feb. 2009.
17.Rosenblatt, F., “Priciples of neurodynamics. perceptrons and the theory of brain mechanisms”, Springer, Berlin, Heidelberg.
18. Geoffrey E. Hinton, Osindero Simon, and Yee-Whye Teh. "A fast learning algorithm for deep belief nets." Neural computation 18.7 (2006): 1527-1554.
19. Bengio, Yoshua, et al. "Greedy layer-wise training of deep networks." Advances in neural information processing systems 19 (2007): 153
20. Hong-Yu Chen, Sin-Horng Chen “Tone Recognition Using MLP and Prosody Model” National Chiao Tung University (2007)
21.Charles Sutton, Andrew McCallum “An Introduction to Conditional Random Fields for Relational Learning” Department of Computer Science University of Massachusetts, USA
22.Ben-Syuan Huang, Chung-Lin Huang “Abnormal Behavior Detection Using CRF Model with Unsupervised Learning” National Tsing Hua University
23.Y.-C Deng, C,-Y Chiang, “A Study on Voice Activation Detection by Using Neural Networks” National Taipei University