This dataset is used to study the commercial EEG wearable devices and their ability to detect drowsiness.
- Our data are acquisited by 5 devices:
- The experiment last about 15 minutes, with 19 participants.
- During the experiment Participants are asked to watching to a video. This video consist of 6 parts, which is shown as follows:
- Relax (1min) -> Watch Video (4min) -> Relax (1min) -> Close Eyes (4min) -> Relax (1min) -> Watch Video (4min)
- Each participant is required to take down their KSS value after they wathch the video.
- When the participants are watching videos, their EEG signals are acquisited by the 5 devices. Each device has 5-7 participants EEG signal acquisited.
- Our data is in
/FAT_WAKE_data - Our Processing code are in
data_process- EEGFileList.py is the file name list
- FW_Class_SVM_KNN.py is KNN and SVM defination
- mlpeeg.py and eegcnn.py is mlp model and cnn model respectively
- feature_selection.py contains the feature selection, F-Score calculation and classify after selection. Change the variables in the first several lines to choose selection options
- FW_Classify.py contains all FAT-WAK experiment data classify. Change the variables in the first few lines to config the classify options.
- TGAM_patern.py contains TGAM Music-Relax classify. Change the variables in the first few lines to config options.
- Our annotations are in
annotation - the methods to connect to devices are in
DeviceConnect - This dataset is also used to teach students in Tongji University, so you may find some tutorial comments.
- Notice: the data are stored using lfs, so when clone please use
git lfs clone https://github.com/YayueHou/FAT-WAKE.git
- We could not promise that data download directly by
Download ZIPis complete