| |
| Gesture recognition with acceleration signal applied to human-computer interaction task, which is continuous signal. There are a number of inter-class variations and intra-class variations, such as time shift and different user operation. Based on these properties, this paper proposes a noval continuous gesture recognition method based on three-axis acceleration sensor, which contains motion segmentation model, removing noises model, normalized features extraction model. Then we constructing supervised classifier using SVM (Support Vector Machines). And we evaluate our method under the scenarios of user-dependent and user-independent, respectively. The experimental results show that our proposed method achieves excellent recognition accuracy under two scenarios. |
|
| Keywords: Pattern Recognition; Continuous Gesture Recognition; Acceleration Sensor; SVM |
|
