Research Article
Motion Intent Recognition in Intelligent Lower Limb Prosthesis Using One-Dimensional Dual-Tree Complex Wavelet Transforms
Table 5
Comparison of our method with other methods with the same data set under user-dependent classification.
| | Subject | Position of sensors | Feature extraction | Classifier | Type of motion state | Accuracy | | Steady | Transitional |
| | Ten able-bodied | Healthy side | Mean, variance, and so on | SVM | 5 | 8 | 95.12% | | Ten able-bodied | Healthy side | — | GMM-HMM | 5 | 8 | 96.92% | | Ten able-bodied | Healthy side | Self-selection feature of CNN | Softmax | 5 | 8 | | | One amputee | | | Ten able-bodied | Healthy side | Five-layer low-frequency coefficients of 1D-DTCWT | SVM | 5 | — | | | — | 8 | | | 5 | 8 | | | One amputee | 5 | — | | | — | 8 | | | 5 | 8 | |
|
|
