Table 1.
Comparative chart of sleep posture recognition.
| S. no | Sensor name | No. of subjects | No. of sensors | Algorithm | Accuracy rate | Year of work |
|---|---|---|---|---|---|---|
| 1 | Uniformly distributed pressure sensors | 10 | 1768 | Support vector machines | 77.14% | 2015 |
| 2 | Uniformly distributed FSR sensor | 19 | 3200 | Deep neural networks | 99.70% | 2018 |
| 3 | Matrix of FSR sensors | 6 | NA | Template matching by a minimum mean squared error | 96.10% | 2017 |
| 4 | Uniformly distributed pressure sensors | 12 | 1728 | Fully connected networks | 97.90% | 2020 |
| 5 | Uniformly distributed pressure sensors | 19 | 3200 | Deep neural networks | 97.10% | 2018 |
| 6 | Uniformly distributed pressure sensors | 14 | 8192 | EMD+K-nearest neighborhood | 91.21% | 2016 |
| 7 | Uniformly distributed pressure sensors | 16 | 1024 | ResNet | 90.08% | 2021 |
| 8 | Accelerometer pulse sensor (our method) | 14 | 5 | Posture prediction-Bayesian network (PP-BN) | 91.05% | 2022 |