2011 Fiscal Year Final Research Report
Study on Automatic Spatiotemporal Labeling Using Adaptive Learning
| Project/Area Number |
21700257
|
|---|
| Research Category |
Grant-in-Aid for Young Scientists (B)
|
| Allocation Type | Single-year Grants |
|---|
| Research Field |
Sensitivity informatics/Soft computing
|
|---|
| Research Institution | Akita Prefectural University |
|---|
Principal Investigator |
|
|---|
| Project Period (FY) |
2009 – 2011
|
| Keywords | ソフトコンピューティング / 適応共鳴理論 / 自己組織化マップ |
|---|
| Research Abstract |
The aim of this study was to develop a method to create category maps using Adaptive Resonance Theory(ART) and Self-Organizing Maps(SOM). As a classification problem that is unknown the number of categories, we evaluated our method to classify of facial expression images. We developed an original dataset of facial expression images consisted of 10 subjects during twenty weeks in one week interval. For the evaluation experiments, our method was able to describe the relations between spatial distribution of and time-series features on dynamic diversity of facial expressions. Moreover, we applied our method to generic object recognition on robot vision to show the extensity of category maps.
|
