Due to the low efficiency of the email application, we are pleased to announce that FatigueView is ready to download directly. You can download the original videos and corresponding annotations (JSON files) directly from https://github.com/FatigueView/fatigueview/tree/main/dataset. Along with the approval of licenses, we will release (publicly available) the rest annotations one by one at this address.
FatigueTree is ready to download from https://github.com/FatigueView/fatiguetree. It contains demo videos of various visual drowsiness features.
All the source codes of our baselines are detailed here. For methods without souces codes (most of them actually), we reproduced them based on the descriptions in their original paper. If authors find their methods are twisted negatively, please contact us.
This is motivated by the fact that most methods evaluated in our article don't have public source codes.
If you find this project useful for your research, please use the following BibTeX entry.
@article{Yang2022FatigueView,
author = {Yang, Cong and Yang, Zhenyu and Li, Weiyu and See, John},
journal = {IEEE Transactions on Intelligent Transportation Systems}
title = {FatigueView: A Multi-Camera Video Dataset for Vision-based Drowsiness Detection},
year = {2023},
volume = {24},
number = {1},
pages = {233-246},
doi = {10.1109/TITS.2022.3216017},
}