Table 1 Some of existing algorithms employed for objective mental health metrics evaluation.
From: Identifying mental health status using deep neural network trained by visual metrics
Author | Algorithm | Accuracy | Data | Task | Subjects | Mental health metric | Number of classes |
|---|---|---|---|---|---|---|---|
Yamada et al.37 | Support vector machine | 91% | Visual features | Watching video | 18 | Mental fatigue | 2 |
Zhai et al.67 | Naive Bayes algorithm, decision tree classifier, and SVM | 90% | Eye gaze, skin temperature, blood volume | Computer game | 32 | Stress | 2 |
Wu et al.39 | Naive Bayes algorithm | 85% | Eye-tracking measures | Robotic skills simulation tasks | 8 | Mental workload | 2 |
Alghowinem et al.68 | Feature fusion + Support Vector Machine | 88% | Speaking behavior, eye activity, and head pose | Verbal and non-verbal behavior | 60 | Depression | 2 |
