1987 Fiscal Year Final Research Report Summary
Application of high Quality Speech Synthesizer to Tutoring Instlement on Lingustics
| Project/Area Number |
61850072
|
|---|
| Research Category |
Grant-in-Aid for Developmental Scientific Research
|
| Allocation Type | Single-year Grants |
|---|
| Research Field |
電子機器工学
|
|---|
| Research Institution | Osaka University |
|---|
Principal Investigator |
KAKUSHO Osamu I.S.I.R., Osaka University Professor, 産業科学研究所, 教授 (30029839)
|
|---|
| Co-Investigator(Kenkyū-buntansha) |
山村 彰 松下電工株式会社, 総合技術研究所, 技手
YAMASHITA Youichi I.S.I.R., Osaka University Techinical Official, 産業科研究所, 教務職員 (80174689)
MIZOGUCHI Riichiro I.S.I.R., Osaka University Associate Professor, 産業科学研究所, 助教授 (20116106)
YANAGIDA Masuzo Ministry of Posts and Telecomunication Chief, 産業科学研究所, 助教授 (00116120)
YAMAMURA Akira Matsushita Electric Works, Ltd. Engineer
|
|---|
| Project Period (FY) |
1986 – 1987
|
| Keywords | Speech / Speech Synthesis / Speech Coding / PARCOR / Bar--code / Speech Dictonary / Turoring Instrument / 語学 |
|---|
| Research Abstract |
This project aims at developring a sppech dictionary which involves bar-codes in terms of phonetic teansctiptions. The usets can hear the synthetis sounds by means of tracing the printed bar-codes with the bar-code reader. The final aim of this porject is that the speech dictionary for japanese students who lean Englist is debveloped. However, in order that investigators involved int this project can evalaute the quality of synthetic sounds. the target language is fixed to Japandes. If speech segments ans synthesis ryles are replaced with those for English, English texts san be also synthesized. The follwing results are derived through this project.
1. The controlling of prosodic parameters is very importnat to realize the high queality speech synthesizer. It is effective that prosodic parameters are controlled by using the imformation extarcted from the natural utterance.
2. The system fro printing speech bar-codes is developed on the personal computer.
3. The trial production consists of two moduls. One generates bar-codes from speech wave. anothor converts bva-codes into synthetic sounds.
4. Synthetic sounds are generated by concatenating several speech segments. The quality of synthetic sounds is fairly good within continuous speech synthesis for Japanese.
5. The method of speech analysis is improved to prepare the better quality of speech segments.
|
