| Project/Area Number |
11CE2005
|
|---|
| Research Category |
Grant-in-Aid for COE Research
|
| Allocation Type | Single-year Grants |
|---|
| Research Institution | Nagoya University |
|---|
Principal Investigator |
ITAKURA Fumitada ITAKURA,Fumitada, 工学研究科, 教授 (30168299)
|
|---|
| Co-Investigator(Kenkyū-buntansha) |
TOYAMA Katsuhiko TOYAMA,Katsuhiko, 情報科学研究科, 助教授 (70217561)
ONISHI Noboru ONISHI,Noboru, 情報科学研究科, 教授 (70185338)
KAKEHI Kazuhiko KAKEHI,Kazuhiko, 情報科学研究科, 教授 (90262930)
TAKEDA Kazuya TAKEDA,Kazuya, 情報科学研究科, 教授 (20273295)
|
|---|
| Project Period (FY) |
1999 – 2003
|
| Project Status |
Completed (Fiscal Year 2003)
|
| Budget Amount *help |
¥907,830,000 (Direct Cost: ¥839,100,000、Indirect Cost: ¥68,730,000)
Fiscal Year 2003: ¥143,000,000 (Direct Cost: ¥110,000,000、Indirect Cost: ¥33,000,000)
Fiscal Year 2002: ¥154,830,000 (Direct Cost: ¥119,100,000、Indirect Cost: ¥35,730,000)
Fiscal Year 2001: ¥170,000,000 (Direct Cost: ¥170,000,000)
Fiscal Year 2000: ¥180,000,000 (Direct Cost: ¥180,000,000)
Fiscal Year 1999: ¥260,000,000 (Direct Cost: ¥260,000,000)
|
|---|
| Keywords | Blind signal separation / Audio-Visual signal processing / Speech recognition / Speech enhancement / Multiple microphones / Spoken language processing / Sound perception / Spatial sound reproduction / 音声信号処理 / 空間音響 / 音声情報処理 / 言語情報処理 / 音認知 / 音データベース / 音声対話 |
|---|
| Research Abstract |
1.Many fundamental research projects performed on the interdisciplinary research topics. (1)Spatial aspects and information measure of the sound signals are integrated through the research on blind signal separation and microphone array beam-formers. (2)The sensation and representation of the spatial audio is analyzed from both viewpoints of signal structure and perception. (3)HRTF and speaker array technologies are studied for 3D sound reproduction. (4)Spatial sound acquisition using multiple microphones is applied to speech enhancement and speech recognition (information transformation). (5)Acoustic and language knowledge are integrated through a large spoken dialogue corpus in the in-car speech interface technology. (6)Speech signal processing technology is extended to deal with the human behavioral signals.
2.Real application systems are developed in order to evaluate the effectiveness of fundamental results in a real world. (1)In-car spoken dialogue system was build for restaurant guidance task, used for data collection, and improved so that it can deal with multiple topics in a dialogue. (2)A PDA (Personal Data Assistance) based speech input system is applied to an interface to a robot guidance system. (3)A robot system which can follow the direction of arrival of the sound is developed.
3.A large speech and acoustic corpora of real world acoustic signals is constructed. The largest corpus is of the in-car spoken dialogue which is recorded through 16ch distributed microphones. Since human behavioral signals associated with driving are also recorded, the corpus is recognized as quite unique and important research resources by the research community.
|
