Research Article
Driver Fatigue Detection Based on Facial Key Points and LSTM
Table 6
Fatigue detection accuracy in the self-built dataset.
| Method | Fatigue detection accuracy (%) |
| MTCNN + DLIB | 85 | Head pose estimation [27] | 80 | Viola–Jones [30] | 85 | DLIB + LSTM | 75 | Proposed | 90 |
|
|