查詢結果分析
相關文獻
- Bayesian Network Based Generation of Prosodic Information for Chinese Text-to-Speech Conversion
- A Mandarin Text-to-Speech System
- 電話障礙查修系統之軟體設計
- HMM式中文詞性自動標注系統
- 應用「拜氏網路」於線上中文簽名確認之研究
- 多重專家應用於線上手寫中文文字識別
- 線上手寫中文文字辨認系統的筆段抽取
- The Survey of On-Line Chinese Character Recognition
- On Speeding Radical-Based On-Line Chinese Character Recognition Through the Directional Matching of the First and Last Segments
- 言為心聲:《哈姆雷》劇中柯勞狄的語言及其兩段獨白的中譯
頁籤選單縮合
題 名 | Bayesian Network Based Generation of Prosodic Information for Chinese Text-to-Speech Conversion=中文文句翻語音中以拜氏網路為基礎音韻訊息之產生 |
---|---|
作 者 | 吳宗憲; 陳昭宏; | 書刊名 | Proceedings of the National Science Council : Part A, Physical Science and Engineering |
卷 期 | 21:5 1997.09[民86.09] |
頁 次 | 頁505-512 |
分類號 | 312.23 |
關鍵詞 | 中文; 文句翻語音; 拜氏網路; 音韻訊息; Text-to-speech conversion; Prosodic information; Pitch contour; Bayesian network; |
語 文 | 英文(English) |
中文摘要 | 本論文中, 我們提出了一利用拜氏網路以產生音韻訊息之方法。 本系統採用 1410 個國語單音作為語音合成單元。 為加強傳統方法中利用規則作為音韻訊息調整之方式 ,本系統利用 112 個音韻平衡句及 250 個挑選之文句作為訓練資料庫。音韻訊息包含音高 走勢、音節強度、音節長度及音節間矩。此外我們利用基週同步疊加演算法來調整音高走勢 。在實驗方面,我們測試 20 個聽眾,結果顯示平均可辨度為 97.0 %,而在自然度方面, 平均鑑定分數( MOS )為 3.8 分。 |
英文摘要 | In this paper, a novel approach based on Bayesian networks to the generation of prosodic information is proposed. A set of 1410 Mandarin syllables is adopted as the basic synthesis units. To enhance the traditional rule-based approach to the generation of prosodic information, the Bayesian network is employed to model the relation between the prosodic information and the linguistic features. This network is trained with a set of 112 phonetically balance sentences and 250 sentences selected from newspapers and textbooks. Given a Chinese character sequence, the Bayesian network can provide appropriate prosodic information including pitch contour, syllable intensity, syllable duration and pause duration. Furthermore, pitch contour modification is achieved by modifying the waveform output using the pitch-synchronous overlap-and-add (PSOLA) method. The synthesized speech has been tested on 20 subjects. The results indicated that the average correct rate was 97.0% for intelligibility, and that the mean opinion score (MOS) was 3.8 form naturalness. |
本系統中英文摘要資訊取自各篇刊載內容。