PhyDAA: Physiological Dataset Assessing Attention

Physiological Dataset Assessing Attention (PhyDAA)

This dataset proposes physiological recordings including electroencephalogram (EEG), eye-tracking and head movement during two tasks enhancing attention. 

Dataset summary

The dataset is composed of the recordings of 32 participants during attention-related tasks in virtual reality. For more information about the design of the tasks please refer to Delvigne V., Ris L., Dutoit T., Wannous H. and Vandeborre J-P., "VERA: Virtual Environments Recording Attention", IEEE SeGAH 2021.
For each participant, physiological signals labelled with the result of the task (representing attention state between distracted and focused) of each trial are provided. 

Dataset structure

The dataset includes:

The raw signals recordings before the pre-processing and feature extraction step in .vhdr (brainvision files) for EEG and .txt for physiological recordings. 

The preprocessed and cut into segments EEG with their corresponding temporal, frequential and handcrafted features. Moreover, two representation are given for features: in array or image form (see original paper for further details).

The scripts used to preprocess and extract the features from raw EEG files. 

Moreover, all the proposed models and a notebook to get started with the files is provided on Github.

Citation

If you use the dataset, please cite the original paper:

@article{delvigne_phydaa_2021,
title = {{PhyDAA}: {Physiological} {Dataset} {Assessing} {Attention}},
issn = {1558-2205},
shorttitle = {{PhyDAA}},
doi = {10.1109/TCSVT.2021.3061719},
journal = {IEEE Transactions on Circuits and Systems for Video Technology},
author = {Delvigne, V. and Wannous, H. and Dutoit, T. and Ris, L. and Vandeborre, J.-P.},
year = {2021},
pages = {1--1}
}

Further information

For further information or questions about the dataset or attention-related tasks please feel free to contact us at [e-mail]. You can also refer to the previously mentioned publication.