Project/Area Number |
22500202
|
Research Category |
Grant-in-Aid for Scientific Research (C)
|
Allocation Type | Single-year Grants |
Section | 一般 |
Research Field |
Sensitivity informatics/Soft computing
|
Research Institution | The University of Tokushima |
Principal Investigator |
FUKUMI MINORU 徳島大学, ソシオテクノサイエンス研究部, 教授 (80199265)
|
Project Period (FY) |
2010-10-20 – 2014-03-31
|
Project Status |
Completed (Fiscal Year 2013)
|
Budget Amount *help |
¥2,860,000 (Direct Cost: ¥2,200,000、Indirect Cost: ¥660,000)
Fiscal Year 2012: ¥780,000 (Direct Cost: ¥600,000、Indirect Cost: ¥180,000)
Fiscal Year 2011: ¥910,000 (Direct Cost: ¥700,000、Indirect Cost: ¥210,000)
Fiscal Year 2010: ¥1,170,000 (Direct Cost: ¥900,000、Indirect Cost: ¥270,000)
|
Keywords | 筋肉電位(EMG) / 統計学習法 / Simple-FLDA / 筋肉電位 / 統計学習 / 手首足首 / オンライン学習 / カーネル関数 / Simple-FDA / EMG識別 / ジャンケン認識 / 手首 / 足首 / Simle-FLDA / Simle-PCA / 統計的学習法 / 筋電 / 近似的学習 / パターン認識 |
Research Abstract |
In this research, first, a simple Kernel Discriminant Analysis (Simple-FLDA) for higher recognition performance by applying a kernel trick to the linear Simple-FLDA (Fisher Linear Discriminant Analysis) is developed. The Simple-FLDA algorithm is composed with simple calculations, but in recognition experiments using the UCI datasets as well as face image dataset; its features equal or surpass those of the Simple-FLDA algorithm. Next, finger motions of the Janken "rock", "Scissors", "paper" and when not doing anything "neutral" were recognized by using wrist EMG signals. Off-line recognition and On-line recognition accuracies are 80% and 70%, respectively. Furthermore, three ankle motions were recognized by using ankle EMG signals. In this experiment, feature vectors are generated by On-line training using the Simple FLDA. Recognition accuracy is greater than 95%.
|