| Project/Area Number |
11650406
|
|---|
| Research Category |
Grant-in-Aid for Scientific Research (C)
|
| Allocation Type | Single-year Grants |
|---|
| Section | 一般 |
|---|
| Research Field |
情報通信工学
|
|---|
| Research Institution | National Institute Of Multimedia Education |
|---|
Principal Investigator |
YUKI Kiyohiro National Institute Of Multimedia Education, Research & Development, Professor, 研究開発部, 教授 (60300696)
|
|---|
| Co-Investigator(Kenkyū-buntansha) |
ASAI Kikuo National Institute Of Multimedia Education, Research & Development, Associate Professor, 研究開発部, 助教授 (90290874)
OSAWA Noritaka National Institute Of Multimedia Education, Research & Development, Associate Professor, 研究開発部, 助教授 (30251721)
SUGIMOTO Yuji National Institute Of Multimedia Education, Research & Development, Professor, 研究開発部, 教授 (90311167)
KONDO Kimio National Institute Of Multimedia Education, Research & Development, Professor, 研究開発部, 教授 (40249925)
YASUDA Hiroshi University of Tokyo, Professor, 先端技術科学研究センター, 教授 (10292765)
|
|---|
| Project Period (FY) |
1999 – 2001
|
| Project Status |
Completed (Fiscal Year 2001)
|
| Budget Amount *help |
¥3,100,000 (Direct Cost: ¥3,100,000)
Fiscal Year 2001: ¥1,100,000 (Direct Cost: ¥1,100,000)
Fiscal Year 2000: ¥1,000,000 (Direct Cost: ¥1,000,000)
Fiscal Year 1999: ¥1,000,000 (Direct Cost: ¥1,000,000)
|
|---|
| Keywords | Virtual Environment / Multi-site Conference / Multi-channel Echo Canceller / Adaptive Algorithm / Variable Step Gain |
|---|
| Research Abstract |
The research purpose is to develop multi-channel voice control system with an echo canceling function, so that one can enhance immersion in the sound environment of multi-site videoconference. To avoid concentration and complexity of the voice arrangement, voice control systems should be put in each station. We studied on architecture and fast algorithm of a digital adaptive filter that remakes echo sounds even with sound resource motion.
We used the FIR filter and Affine projection algorithm with variable step gain. The simulation result shows effectiveness of the method that the step gain is modified with voice existence. Then, we proposed a method that used volume information in the adaptive algorithm of an echo canceller. The method estimated echoes in the two-channel echo canceller within time that does not give conversation much interfere.
In the remote conference, it is considered remote speakers talk simultaneously. Then, we simulated the situation that had one input with voice and another with noise. The result showed an echo path in the voice channel was estimated well, and estimation error in the noise channel saturated. In the noise channel, the step gain became small because of small input signal, acid the filter coefficients were not modified mostly.
|
