Deep Convolutional Network Approach in Spike Train Analysis of Physiotherapy Movements
Classifying gestures or movements nowadays have become a demanding business as the technologies of sensors have risen. This has enchanted many researchers to actively and widely investigate within the area of computer vision. Physiotherapy is an action or movement in restoring someone’s to health wh...
Saved in:
| Main Authors: | , , , , , |
|---|---|
| Other Authors: | |
| Format: | Book Chapter |
| Language: | en |
| Published: |
Springer, Singapore
2019
|
| Subjects: | |
| Online Access: | http://ir.unimas.my/id/eprint/40404/3/Deep%20Convolutional%20-%20Copy.pdf http://ir.unimas.my/id/eprint/40404/ https://link.springer.com/chapter/10.1007/978-981-15-1289-6_15 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Classifying gestures or movements nowadays have become a demanding business as the technologies of sensors have risen. This has enchanted many researchers to actively and widely investigate within the area of computer vision. Physiotherapy is an action or movement in restoring someone’s to health where they need continuous sessions for a period of time in order to gain back the ability to cope with daily living tasks. The rehabilitation sessions basically need to be monitored as it is essential to not just keep on track with the patients’ progression, but as well as verifying the correctness of the exercises being performed by the patients. Therefore, this research intended to classify different types of exercises by implementing spike train features into deep learning. This work adopted a dataset from UI-PRMD that was assembled from 10 rehabilitation movements. The data has been encoded into spike trains for spike patterns analysis. Spike train is the foremost choice as features that are hugely rewarding towards deep learning as they can visually differentiate each of the physiotherapy movements with their unique patterns. Deep Convolutional Network then takes place for classification to improve the validity and robustness of the whole model. The result found that the proposed model achieved 0.77 accuracy, which presumed to be a better result in the future. |
|---|