Desigining Privacy-Aware Collaboration Tool with Eye-tracking Technology
With our current technology, it is often more practical and time-saving to conduct meetings as virtual online meetings over the Internet. Compared to traditional telephone-based conference calls it is now also possible to even hold presentations online and present slides to the listeners. A disadvantage of online presentations is that the presenter has no way of knowing whether the listeners are looking at the right part of the slides. This is a problem when explaining complicated slides, e.g., graph plots, especially if multiple plots are on the same slide. Giving the presenter feedback about the current attention of its listeners can also be useful to speed up or slow down the presentation when the listeners are already reading ahead or are still stuck on previous parts of the slides, respectively. Providing the listeners with information about the presenters attention can be advantageous as well, since it allows the listeners to see which part of the presentation the presenter is currently looking at.
This feedback can for example be provided by tracking the eye movement of a user, calculating what the user is looking at and transmitting this information to the other users. A problem with transmitting this information directly is that this allows to find out which individual listener is looking at which part of the presentation. While being a privacy invasion in general this could also be used to discriminate inattentive employees in a business setting.
In this thesis an online presentation system should be designed and implemented that allows the listeners and presenters to see where the current attention is at, while still preserving the privacy of the users. Specifically this means that it should be impossible for anyone to find out which of the listeners is looking at which part of the presentation To improve the trustworthiness of the system, the privacy preserving properties of the system should be made clear to its users.
The resulting implementation is planned to be presented at ZKM.
A functioning eye tracking solution will be provided
Programming skills (Java; C\# or Python; networking) are required.