2012 Fiscal Year Final Research Report
Research about source separation and source localization bynonlinear array signal processing based on quadratic kernel
Project/Area Number |
23760325
|
Research Category |
Grant-in-Aid for Young Scientists (B)
|
Allocation Type | Multi-year Fund |
Research Field |
Communication/Network engineering
|
Research Institution | University of Tsukuba |
Principal Investigator |
MIYABE Shigeki 筑波大学, システム情報系, 助教 (50598745)
|
Project Period (FY) |
2011 – 2012
|
Keywords | パターン認識 |
Research Abstract |
We studied source separation and source localization by nonlinear microphone array signal processing using higher dimensional map with kernels, and successfully developed high-resolution source localization using quadratic or higher-degree mapping. We showed that the covariance analysis of the mapping corresponds to the moment analysis with the order twice of the degree of the map, and the statistical property of the analysis in the source localization is clarified. It is ascertained that the proposed source localization has higher resolution and is more computationally efficient than the conventional methods.
|