2014 Fiscal Year Final Research Report
Generative model in a wide class of distribution and its application
| Project/Area Number |
24500165
|
|---|
| Research Category |
Grant-in-Aid for Scientific Research (C)
|
| Allocation Type | Multi-year Fund |
|---|
| Section | 一般 |
|---|
| Research Field |
Intelligent informatics
|
|---|
| Research Institution | The University of Electro-Communications |
|---|
Principal Investigator |
|
|---|
| Co-Investigator(Kenkyū-buntansha) |
HOTTA Kazuhiro 名城大学, 理工学部, 准教授 (40345426)
|
|---|
| Project Period (FY) |
2012-04-01 – 2015-03-31
|
| Keywords | 自己相関カーネル / サポートベクトルマシン / Ｆｉｓｈｅｒカーネル / 顔検出 |
|---|
| Outline of Final Research Achievements |
We studied learning machine using the generative models and simple discriminator. We proposed a new framework of Markov random field that has kernelalized potential function. We showed an efficient method of computation and that this model generates essentially linearly separable kernel features if the degree of kernel is large.
We conducted experiments on the face detection using the appearance based approach, and showed that our method can attain comparable results with the state-of-the-art face detection methods based on AdaBoost, SURF, and cascade despite of smaller data size and no preprocessing.
|
| Free Research Field |
情報通信（機械学習とパターン認識）
|
