| Project/Area Number |
13650461
|
|---|
| Research Category |
Grant-in-Aid for Scientific Research (C)
|
| Allocation Type | Single-year Grants |
|---|
| Section | 一般 |
|---|
| Research Field |
Measurement engineering
|
|---|
| Research Institution | Kobe University |
|---|
Principal Investigator |
KOTANI Manabu Kobe University, Faculty of Engineering, Associate Professor, 工学部, 助教授 (30215272)
|
|---|
| Co-Investigator(Kenkyū-buntansha) |
OZAWA Seiichi Kobe University, Graduate School of Science and Technology, Associate Professor, 大学院・自然科学研究科, 助教授 (70214129)
|
|---|
| Project Period (FY) |
2001 – 2002
|
| Project Status |
Completed (Fiscal Year 2002)
|
| Budget Amount *help |
¥1,900,000 (Direct Cost: ¥1,900,000)
Fiscal Year 2002: ¥700,000 (Direct Cost: ¥700,000)
Fiscal Year 2001: ¥1,200,000 (Direct Cost: ¥1,200,000)
|
|---|
| Keywords | Independent Component Analysis / Pattern Recognition / Signal Processing / Neural Networks / Feature Extraction / Biobgical Signal Processing / EMG / Acoustic Diagnosis / ニュートラルネットワーク |
|---|
| Research Abstract |
We examine an application of independent component analysis (ICA) to feature extraction of signal processing such as digit patterns and acoustic signals. In order to evaluate the effectiveness of independent components as features, we compare discrimination accuracy using independent components with those using principal components. Furthermore, we apply the ICA to biological signal processing. We obtain the following results :
1. Acoustic diagnosis
In order to detect the leakage from pipesaccurately, we should select a feature extraction method for sounds properly. The purpose of this research is to examine whether independent component analysis (ICA) is useful as a feature extraction method for acoustic signals. We confirm that the feature extraction using the ICA algorithm is very useful for detecting gas leakage sounds.
2. Digit recognition
We propose a novel recognition method using features extracted by ICA. The proposed method consists of some modules for each category and a synthesizer. We evaluate the performance of the proposed method for several recognition tasks. From these results, we confirm the effectiveness of the recognition method using independent components for each class.
3. Deconvolution for EMG
We apply a multichannel blind deconvolution method based on ICA to surface EMG signals. We obtained a few components of which firing patterns is similar to motor units.
|
