Markerless 3D hand-tracking for unencumbered human-computer interaction
Hand gestures play an important role in human-to-human interaction and ideally, we should be able to use them to interact with computers in an equally seamless and natural manner. Vision-based tracking has great potential as an enabling technology for unencumbered 3D human-computer interaction because it offers a cost-effective, non-invasive alternative to expensive hardware devices such as gloves, 3D mice and motion capture systems.
The aim of this research is to develop new methods for tracking and recovering the 3D pose of hands (in real-time) without the use of special markers. In particular, this project will investigate the tracking of motions involving two hands using either monocular, multiple views or depth camera , with emphasis on modelling high-level interactive behaviours between the two hands.
Candidates should have (or expected to obtain) a first class degree and/or Masters with distinction in a numerate discipline with good ability in mathematics and excellent programming skills in C/C++ and/or MATLAB. An interest in computer vision is essential and relevant experience an advantage.