Help-Me-Watch provides video content tailored to each student.
The app aims to take some of the stress out of exam revision for thousands of students at DCU by enabling them to hone in quickly and easily on specific segments of their online lectures and tutorials that they may have missed.
Help-Me-Watch was created by a team of researchers at DCU’s Faculty of Engineering and Computing and the Insight SFI Research Centre for Data Analytics and is available to DCU lecturers and students.
Various distractions
The app was developed as a response to student feedback on their experiences with online lectures and tutorials after the arrival of COVID-19.
Students highlighted that they found it challenging to remain attentive during online learning sessions because of various distractions, citing examples such as sharing a room with others, general noise within the home with family members working from home or other examples from answering the door to take deliveries to dogs barking.
The system provides support for student learning in online, synchronous zoom lectures, tutorials, workshops and lab sessions. It tailors video recordings of those sessions so that each student gets a different video summary, personalised to their individual attention levels.
“Early findings on the use of the app are that students find it will likely be of most benefit later in the semester when it comes to revision for their examinations," said Prof Alan Smeaton.
"Students, owing to the pivot to online learning because of COVID-19 may be faced with dozens of hours of video recordings of live sessions for each of their modules.
"Help-Me-Watch will help them find the crucial parts in those video recordings that they may have missed and may help to take some of the stress out of exam revision.”
Artificial intelligence techniques
Outlining the nature of the system, Prof Smeaton explained that Help-Me-Watch uses AI (artificial intelligence) techniques to record student attention during the online synchronous zoom session. It then generates video summaries of the live session, personalised and tailored for each student.
These summaries are then used when a student is re-capping material, or when revising for an exam and allows more targeted playback of video recordings.
“The system runs through a web browser on a Mac or Windows laptop or desktop, is GDPR compliant and completely anonymises student identities so we never know which students are watching,” added Prof Smeaton.
The research team including lecturers in Computing and Electronic Engineering Hyowon Lee, Mingming Liu and Michael Scriney, will present findings and observations on Help-Me-Watch at the upcoming virtual Association for the Advancement of Artificial Intelligence conference in February.