Table 2.
Classification accuracy of shoulder postures using the proposed convolutional neural network (CNN) networks (%).
| Posture | Subject | Mean | ||||
|---|---|---|---|---|---|---|
| 01 | 02 | 03 | 04 | 05 | ||
| Rest | 98.00 | 98.00 | 98.60 | 98.20 | 99.20 | 98.40 |
| Upper Rotation (Elevation) | 96.60 | 94.80 | 96.00 | 97.40 | 94.40 | 95.84 |
| Depression | 98.40 | 96.20 | 99.00 | 97.20 | 97.20 | 97.60 |
| Protraction | 97.80 | 96.00 | 97.20 | 97.60 | 97.00 | 97.12 |
| Retraction | 98.00 | 96.60 | 97.20 | 96.60 | 96.00 | 96.88 |
| Mean | 97.76 | 96.32 | 97.60 | 97.40 | 96.76 | 97.17 |