Posting a little update on the development of the next version. During my vacation home to Sweden I visited the ITU GazeGroup in Copenhagen to discuss implementation, pontential applications, hardware and new areas of research. The ITU group is now located on the second floor of the brand new Mikado building which exemplifies modern Danish architecture.  I met up with Doctor Javier San Agustin to continue development of the big 2.0 version and brought a number of cameras, lenses, LEDs and filters to experiment with.

There are a lot of things to discuss when building a new version that is capable of several methods ofeye tracking, such as head mounted or remote from one unified platform. During a week we spent all day discussing, sketching and implementing. Our primary goal was to incorporate remote eye tracking, a feature that was experimental in version 1.0 and a current requests amongst all users. Another item for re-design was the user interface. As we are adding more options to the platform the interaction needs to be streamlined and simplified. We decided to seperate the configuration panel into basic and advanced settings which by default are hidden but easily are reachable for more advanced users.

Since the gathering we have released the software to the public and made several important improvements. Remote tracking is now fairly stable and performance have reached an unparalleled level in the open source domain. Full binocular tracking is now available at 120Hz or more. Basic test demonstrates monocular tracking up to 300Hz. On a modern computer the major limiting factor is the number of frames per second the camera can provide.

During my visit I also had the opportunity to meet the group members, John Paulin Hansen who has extensive experience with gaze interaction and a major organizer of the COGAIN association. Henrik Skovsgaard, phd student, who helped with initial development of the DirectShow camera class and bugtesting based on the posts in the forum. Hopefully the next version will be free of the black-image issue. I also met with David Rozado (drozado on the forum) who is visiting the group to work on a new exciting project on gaze gesture recognition using HTM neural network algorithms. The ITU group has also a new PhD student, Diako Mardanbeigi, who demonstrated an impressive masters thesis last year. Really nice to catch up with fellow eye tracking people.