Dissertation > Excellent graduate degree dissertation topics show

Study on the Key Technology of Tibetan Speech Synthesis

Author: SuoNanZhaXi
Tutor: OuZhu
School: Tibet University
Course: Chinese Ethnic Language and Literature
Keywords: Speech synthesis The tibetan word segmentation algorithm Marked rhythm Latin conversion Question sets
CLC: H214
Type: Master's thesis
Year: 2011
Downloads: 21
Quote: 0
Read: Download Dissertation

Abstract


Speech synthesis system consists of three main components: text analysis module, the prosody generation module and the acoustic module. This paper, in three aspects of Speech synthesis system traced above, the text preprocessing, respectively studied Tibetan word segmentation, prosodic labeling rules, the rhythm marked content, marked rhythm required Tibetan--Latin conversion algorithm, voice processing back-end synthesis required the design of problem sets. The main work of this paper is as follows:1. This paper does more in-depth research on the various problems of the case particle used in Tibetan words. Tibetan case particle is a more obvious word boundary, as boundary markers, they are conducive to Tibetan word segmentation. The phenomenon of case particle which are out of "eight Case" gets formalized description and study and it provides an effective strategies and approaches for further studying Tibetan word segmentation.2. The core of Speech synthesis will mainly be reflected by text analysis and rhythm control, in which text analysis is the basis for speech synthesis, which includes conversion of special symbols and word cut grading of text. The accuracy of automatic segmentation will seriously affect the naturalness and clarity of the output audio stream and it is the core of text analysis system. Therefore, this paper did a systematic study on Tibetan word segmentation algorithms, and also achieved forward and reverse maximum matching Tibetan word segmentation algorithm. Thanks to the two algorithms do not have the ability to identify unknown words, i proposed Tibetan word segmentation algorithm of labeling and normalization and it has a good recognition of unknown words.3. The speech corpus building occupies a very important position in the Corpus-based speech synthesis technology, it is basis and key of speech synthesis system, in the aspect of verbal mark, the paper developed the corresponding corpus tagging rules while building speech corpus, and also used Latin transliteration for the label. After the achievement of Latin-- Tibetan translation algorithm, the algorithm has a feature of good move set and low complexity.4. Based on the language of Lhasa, the paper designed context attributes (it needs to consider generally the changing characteristics of all parameters), determined the factors need to be considered for Tibetan speech synthesis question sets, it laid good foundation. for achievement of the Tibetan language speech synthesis system.

Related Dissertations

  1. Application of Military Speech Name Database by VTFR Synthetic Method,TN912.3
  2. Mobile audio and video interactive platform of business execution,TN915.09
  3. Robot Control System Simulation,TP242
  4. Applied Research on Dam Monitoring System Based on the Speech Synthesis of Automatic Segmentation and PSOLA,TP277
  5. Ant Application in the Testing for IVR System,TP311.52
  6. Research and Implemenation of Voice Intelligent plantform Based on VoiceXML,TP311.52
  7. Study on Automatic Construction of Speech Database~2,TN912.33
  8. Based on the HMM training Vietnamese speech synthesis system,TN912.33
  9. The Study of Pitch Shifting Algorithms and the Application in Speech Synthesis,TN912.33
  10. Research and development of radio frequency identification information terminal based on embedded technology,TP368.12
  11. With Speech Synthesis Function Based on GSM Network System Design and Implementation,TN912.33
  12. The Analysis of Recommendation G.729 PinYin Synthesis Algorithm and Its Implement Based on DSP,TN912.3
  13. Research on Emotion Speech in Cross-clutural Backgrounds,TP391.42
  14. Based on the emotional pitch template emotional speech synthesis,TN912.33
  15. Design and implementation of speech synthesis software -oriented the meteorological voice call center,TN912.33
  16. . NET-based personalized research and implementation of e-commerce platform,TP311.52
  17. Auto-constructing Speech Corpus with the Limited Text~2,TP391.1
  18. Design and Realization of Text to Speech Communication Systems Based on Web,TP311.52
  19. Speech Synthesis Algorithm Research and FPGA Implementation,TN912.33
  20. Intelligent voice broadcast system design and implementation,TN912.3
  21. Research on Encoding and Recognition of Two-Dimensional Barcode and Distributed Voice Broadcast System,TP391.44

CLC: > Language, writing > China 's minority languages > Tibetan language
© 2012 www.DissertationTopic.Net  Mobile