ETH Zurich : Computer Science : Pervasive Computing : Distributed Systems : Education : Student Projects : Abstract

Wearable Eye Tracking with Mobile Devices (B)

Status: Abgeschlossen

Recent technological advancements have made wearable eye trackers (e.g., Pupil) more affordable and widely accessible. One limitation common to many such platforms is that they do not have any own computing capabilities. They need to be connected to a powerful computer for data processing. Recently, there have been attempts to increase the portability of the Pupil platform. A companion Android application allows recording eye tracking sessions and processing the data offline or streaming video data over the network. While this may be sufficient for statistical analysis of the data, this mode of operation does not support online interactive gaze-aware systems.

To overcome this issue, we are developing an open-source platform to enable real-time wearable eye tracking. Users can connect the Pupil eye tracker to their own Android device and rapidly develop their own prototypes. All the routines required for pervasive eye tracking like pupil detection and calibration are running locally on their own mobile device.

Requirements: solid C++ programming skills, background in image processing is necessary.

Keywords: eye-tracking, image processing, wearables