Thesis Defence: Detecting levels of attention from Electroencephalography (EEG) signals
December 17 at 10:00 am - 2:00 pm
Thuppahiralalage Eranga De Saa, supervised by Dr. Pourang Irani, will defend their thesis titled “Detecting levels of attention from Electroencephalography (EEG) signals” in partial fulfillment of the requirements for the degree of Master of Science in Computer Science.
An abstract for Thuppahiralalage Eranga De Saa’s thesis is included below.
Defences are open to all members of the campus community as well as the general public. Registration is not required for in-person defences.
ABSTRACT
Detecting attention lapses during educational activities, particularly in the context of reading amid auditory distractions, is essential yet often overlooked in current research. In this thesis, we leveraged EEG signals to detect attentional lapses. The participants completed a reading task with and without auditory oddball distractions to establish two distinct attention levels which were validated through NASA TLX scores and reading comprehension outcomes. We compared three feature extraction techniques powerband, Common Spatial Patterns (CSP), and filterbank CSP across five classifiers.
Filterbank CSP achieved the highest accuracy (over 90%), with the beta band outperforming theta and alpha in classification tasks. Notably, classifier choice was less influential with CSP methods. Our reduced-channel classification results suggest the feasibility of using limited-channel EEG
devices for future studies.
