2011 Fiscal Year Final Research Report
Study on a speech understanding system with an ability to estimate the effect of various environments and users on recognition accuracy
| Project/Area Number |
21500165
|
|---|
| Research Category |
Grant-in-Aid for Scientific Research (C)
|
| Allocation Type | Single-year Grants |
|---|
| Section | 一般 |
|---|
| Research Field |
Perception information processing/Intelligent robotics
|
|---|
| Research Institution | Shizuoka University |
|---|
Principal Investigator |
KAI Atsuhiko 静岡大学, 工学部, 准教授 (60283496)
|
|---|
| Co-Investigator(Kenkyū-buntansha) |
KOGURE Satoru 静岡大学, 情報学部, 講師 (40359758)
WANG Longbiao 静岡大学, 工学部, 助教 (30510458)
|
|---|
| Project Period (FY) |
2009 – 2011
|
| Keywords | 音声情報処理 |
|---|
| Research Abstract |
Designing a spoken dialogue interface involves an appropriate handling of recognition errors, which often caused by background noise or indistinct voice uttered by users, and has great impact on the usability of such interface system. This study has developed the methods for estimating user's recognition accuracy by his/her utterance, and also investigated the method to apply the recognition accuracy estimate, which depends on a dialogue state, for optimal selection of responses to the user. Evaluation experiments showed the effectiveness of the proposed methods.
|
