| Project/Area Number |
25350488
|
|---|
| Research Category |
Grant-in-Aid for Scientific Research (C)
|
| Allocation Type | Multi-year Fund |
|---|
| Section | 一般 |
|---|
| Research Field |
Social systems engineering/Safety system
|
|---|
| Research Institution | National Research Institute of Police Science |
|---|
Principal Investigator |
Osanai Takashi 科学警察研究所, 法科学第四部, 部長 (70392264)
|
|---|
| Research Collaborator |
KAMADA toshiaki
MAKINAE hisanori
AMINO kanae
|
|---|
| Project Period (FY) |
2013-04-01 – 2019-03-31
|
| Project Status |
Completed (Fiscal Year 2018)
|
| Budget Amount *help |
¥4,160,000 (Direct Cost: ¥3,200,000、Indirect Cost: ¥960,000)
Fiscal Year 2017: ¥910,000 (Direct Cost: ¥700,000、Indirect Cost: ¥210,000)
Fiscal Year 2016: ¥780,000 (Direct Cost: ¥600,000、Indirect Cost: ¥180,000)
Fiscal Year 2015: ¥910,000 (Direct Cost: ¥700,000、Indirect Cost: ¥210,000)
Fiscal Year 2014: ¥780,000 (Direct Cost: ¥600,000、Indirect Cost: ¥180,000)
Fiscal Year 2013: ¥780,000 (Direct Cost: ¥600,000、Indirect Cost: ¥180,000)
|
|---|
| Keywords | 話者認識 / 発話様式 / 時期変動 / 特徴量変換 / 犯罪捜査支援 |
|---|
| Outline of Final Research Achievements |
In speaker recognition, it is said that the differences in recording environments, speaking styles, and timing of recording speech samples, are one of the factors that cause deterioration of authentication performance. Therefore, we investigated the speaker recognition method that is robust to such differences. In this research, we used various speech databases that we constructed so far. In addition, we used the Standardization-Normalization Transformation, which was proposed earlier in our research and proved to be effective in improving the speaker recognition performances. Results of the experiments showed that the Standardization-Normalization Transformation is an effective method for conquering the differences in the recorded speech data.
|
| Academic Significance and Societal Importance of the Research Achievements |
非協力的な話者を扱うことが多い法科学分野においては、多様な音声資料に適応できる話者認識に期待する声は大きい。例えば、振り込め詐欺事件において、同一犯による犯行の解明には、犯人の音声を比較して同一話者か否かを判断する必要があるが、それぞれの事件の会話はさまざまである上、関係者を装うなど、話し方も多様となるケースが多い。本研究の成果を利用することで、多様な音声資料であってもそれぞれの事件の犯人の同一性を示すことが期待できる。
|
