In this project we tried to prove, that a web-cam based iris localization system can achieve accuracy compared to the IR based head-mounted devices. We have also proposed a state-of-the-art iris center localization algorithm. The results of the project will be soon published in IET Computer Vision in the article “Webcam-based system for video-oculography”.

Watch how our iris localization works:

The source codes for our iris center localization can be found at:

https://github.com/JacekNaruniec/irislocalization

Project was conducted in a cooperation with Department of Neurology, Faculty of Health Science, Medical University of Warsaw and Polish-Japanese Academy of Information Technology, Warsaw.

The goal of this project is to create algorithms for driving animated faces with our own face. The advantage that we have to the other solutions is that we don’t use any specialized devices (like Kinect sensors or markers), only simple webcam .

People involved:

Marek Kowalski

Goal of this project is to merge results from multiple Kinect v2 sensors to create one, beautiful 3D points cloud. Our software is open source, you can find it on GitHub. For more informations visit Marek Kowalski’s homepage.

People involved:

Marek Kowalski

The goal of this project was to create a reliable facial features detection algorithm. The C++ library, working in Windows, also with the demo using webcam, is available at this page.