Proposal of the facial expression classification method based on the divided hierarchical CNN and the evolutional structural adaptive learning algorithm
| Project/Area Number |
19700226
|
|---|
| Research Category |
Grant-in-Aid for Young Scientists (B)
|
| Allocation Type | Single-year Grants |
|---|
| Research Field |
Sensitivity informatics/Soft computing
|
|---|
| Research Institution | Kisarazu National College of Technology |
|---|
Principal Investigator |
OEDA Shinichi Kisarazu National College of Technology, 情報工学科, 講師 (80390417)
|
|---|
| Project Period (FY) |
2007 – 2009
|
| Project Status |
Completed (Fiscal Year 2009)
|
| Budget Amount *help |
¥2,700,000 (Direct Cost: ¥2,400,000、Indirect Cost: ¥300,000)
Fiscal Year 2009: ¥650,000 (Direct Cost: ¥500,000、Indirect Cost: ¥150,000)
Fiscal Year 2008: ¥650,000 (Direct Cost: ¥500,000、Indirect Cost: ¥150,000)
Fiscal Year 2007: ¥1,400,000 (Direct Cost: ¥1,400,000)
|
|---|
| Keywords | ニューラルネットワーク / Cellular Neural Network / 進化的構造適応学習法 / 砂時計型ニューラルネットワーク / Cellular Neural Network (CNN) |
|---|
| Research Abstract |
In this research, we optimized the sandglass type neural network by the evolutional structural adaptive learning algorithm with GA or IA. As a result, we succeeded in the search of the initial connection weight in neural network. A convergence speed of the optimized weight was faster than the initial weight given to it at random. We solved the local minimal problem of the network inversion. It was applied to the sandglass type neural network. To verify the effectiveness of our proposed method, we applied the evolutional structural adaptive learning algorithm to the divided hierarchical CNN and we classified the facial expression image data.
|
Report
(4 results)
Research Products
(9 results)
