Table 2.
State-of-the-art DFD systems used smartphone-based architecture by machine learning algorithms.
| Cited | Sensors/Parameters | Algorithms | Accuracy | Platforms |
|---|---|---|---|---|
| [156] Garc et al. (2014) | Eye movements and PPG signals | ANN, DBN, SVM, ICA and GA | NA | Android |
| [157] Chang et al. (2012) | ECG, PPG, temperature, heart rate, blood pressure, temperature, speed and PERCLOS | Fuzzy Bayesian framework | NA | Android |
| [158] Xu et al. (2014) | PERCLOS, blink time and blink rate | NN | ACC: 90% | Android |
| [159] Zhang et al. (2014) | EEG, ECG, EOG | ACC: 96.5 | Android | |
| [160] Dasgupta et al. (2018) | PERCLOS, Infrared Light and Microphone | Percentage of eyelid | ACC: 93.33% | Android |
| [161] Zhang et al. (2018) | Steering behavior and heart rate of the driver | Wearing smartwatch and second heart rate | ACC: 94.39% | Android |
| [162] Freidlin et al. (2018) | ECG, EMG and galvanic skin response (GSR) modules and accelerometers, a magnetometer and a gyroscope | NA | NA | IOS & Android |
| [163] Bakar et al. (2015) | PERCLOS and GPS | NA | NA | Android |
| [164] Yin et al. (2017) | EEG and PEN | Fuzzy Entropy and SVM | ACC: 95% | Android |
DBN: Dynamic Bayesian network, ANN: Artificial neural network, ICA: Independent component analysis, ACC: Accuracy, ECG: electroencephalography, EEC: electrocardiography, EOG: electrooculography, SVM: support vector machine, GA: Genetic algorithm, NA: Not applicable, PPG: photoplethysmogram.