Full metadata record
DC Field | Value | Language |
---|---|---|
dc.contributor.author | 洪國興 | en_US |
dc.contributor.author | Kuo-Hsing Hung | en_US |
dc.contributor.author | 陳信宏 | en_US |
dc.contributor.author | Sin-Horng Chen | en_US |
dc.date.accessioned | 2014-12-12T02:31:19Z | - |
dc.date.available | 2014-12-12T02:31:19Z | - |
dc.date.issued | 2006 | en_US |
dc.identifier.uri | http://140.113.39.130/cdrfb3/record/nctu/#GT009213645 | en_US |
dc.identifier.uri | http://hdl.handle.net/11536/70834 | - |
dc.description.abstract | 文字轉語音系統中所使用的語料庫已由小量只提供有限基本音節的方式,演變為以大型語料庫為基礎的語料庫。在本論文中,實作了一套以語料庫為基礎之中文文句翻語音系統,除了整合既有合成系統中之文字分析器、韻律訊息產生器與波形合成器外,我們特地設計一個合成單元的選擇機制來解決以語料庫為基礎的合成系統會遇到的兩大問題,如何有效率地由大量語料中搜尋到可用之合成單元?以及如何挑選一個最佳的合成單元連接方式。本實作中,在不損害音質的前提下,針對搜尋效率改進了前人提出的連續相關比對法,也以更加嚴密的方式重新定義了用來挑選合成單元的cost function。 最後,為了此系統使用上的方便性,我們設計了一套圖形化使用者介面。在此介面上,使用者直接輸入文字,然後可依賴系統自動合成語音,或者以其意願選擇合成單元。 | zh_TW |
dc.description.abstract | Synthesis units in Mandarin Text-to-Speech system have migrated from small inventory of base-syllables to a large corpus. In this thesis, a corpus-based Mandarin Text-to-Speech system is implemented. Besides integrating the present text analyzer and the prosodic information generator, the study emphasizes on designing a unit-selection algorithm to solve the two main problems of searching all possible synthesis unit candidates in a huge corpus efficiently and selecting an appropriate unit sequence. We improve the efficiency of the continuous-correlative comparison method without decreasing the quality of the synthesis waveforms. Also we re-define the cost function used in the unit selection. Finally, for users’ convenience, we design a graphical user interface for this system. The user can directly type his input text, and get the synthesis waveform and some intermediate information automatically. | en_US |
dc.language.iso | zh_TW | en_US |
dc.subject | 語音合成 | zh_TW |
dc.subject | 以語料庫為基礎 | zh_TW |
dc.subject | 合成單元挑選 | zh_TW |
dc.subject | Text-to-Speech | en_US |
dc.subject | Corpus-based | en_US |
dc.subject | Unit Selection | en_US |
dc.title | 以語料庫為基礎之中文文句翻語音系統實現 | zh_TW |
dc.title | An Implementation of Corpus-based Mandarin TTS System | en_US |
dc.type | Thesis | en_US |
dc.contributor.department | 電信工程研究所 | zh_TW |
Appears in Collections: | Thesis |
Files in This Item:
If it is a zip file, please download the file and unzip it, then open index.html in a browser to view the full text content.