QMUL-UT EEG Dataset

This dataset is the result of a cooperation between Queen Mary University and the University of Twente. The dataset was acquired from 17 subjects over two weeks in December 2008 at the University of Twente.

We recorded EEG and physiological signals from each subject as the subject watched video sequences depicting events, followed by either a matching or a non-matching tag. We used 49 videos (7 videos in 7 categories) and each video was shown once with a matching tag and once with a non-matching tag. Videos were displayed using the Presentation software and markers are available within the EEG data signaling the start and end-times of videos and tags.

EEG was recorded on a BioSemi ActiveTwo system for 32 electrodes according to the international 10/20 system. We also recorded EOG, GSR, heart rate, temperature and respiration data. After the main experiment, each subject rated each video in terms of valence and arousal. of the 17 subject, 12 were male. Ages ranged from 19 to 31, with a mean age of 25. All but two subjects were right-handed and all but three subjects viewed the tags in their native language.

In our research on the data, we have shown significant differences in EEG signals between videos with matching and non-matching tags. We are currently working on a single trial analysis. More details on the dataset and our research can be found in our ABCI 2009 paper (available on the publications page).

If you are interested in using this data for your research, please send an email to sander (dot) koelstra [at] elec.qmul.ac.uk. We will make the data available to you after signing an EULA. Any researchers using the dataset in a publication must refer to our ABCI paper (bibtex avaialable at the publications page).