Table 4.
Extraction of visual features used to train and test the network based on online state-of-the-art vision-based datasets.
| Cited | Data Source | Features | Link URL |
|---|---|---|---|
| [11] | NTHU-DDD Dataset | 36 subjects, video: 9.5 h, 5 different classes |
http://cv.cs.nthu.edu.tw/php/callforpaper/datasets/DDD/ |
| [180] | UTA-RLDD dataset | Video—30 h, 3 features: alertness, low vigilance, and drowsiness, frame rate: 30 fps, participant: 60 |
http://vlm1.uta.edu/~athitsos/projects/drowsiness/ |
| [181] | MultiPIE | different subjects, poses, illumination, occlusions, 68 landmark points | https://ibug.doc.ic.ac.uk/resources/facial-point-annotations/ |
| - | Kaggle-distracted drivers | 22,424 images of size (480 × 680), 10 classes | https://www.kaggle.com/c/state-farm-distracted-driver-detection |
| [182] | 3MDAD | 60 subjects, 16 different actions |
https://sites.google.com/site/benkhalifaanouar1/6-datasets#h.nzos3chrzmb2 |
| [183] | MiraclHB | AVI format with a resolution of 640 × 480 and frequency 30 fps, 12: subjects | http://www.belhassen-akrout.com/ |
| [184] | BU-3DFE | 100: subjects with 2500 facial expression models | http://www.cs.binghamton.edu/~lijun/Research/3DFE/3DFE_Analysis.html |
University of Texas at Arlington Real-Life Drowsiness Dataset (UTA–RLDD), National Tsing Hua University Drowsy Driver Detection (NTHU–DDD), multiview points, illumination and expressions (MultiPIE), multimodal multiview and multispectral driver action dataset (3MDAD), Multimedia Information Systems and Advanced Computing Laboratory Hypo-vigilance database (MiraclHB), and Binghamton University 3D facial expression (BU–3DFE).