Project Details:
Eye Movement Tracking
Guidance: Prof. M. Hanmandlu
Current Research Project
Executive Summary:
The project is to develop an Efficient Recognition Technique using Eye
Movement Tracking. The most popular biometric methods like fingerprint verification or iris recognition are based mostly on physiological properties of the human body, which can be easily forged. Therefore, for proper identification what is needed is only a “body” of a person which is to be identified. It makes possible to identify an unconscious or - in some methods - even a dead person. Eye movement tracking compiles both physiological (muscles) and behavioral (bandrain) aspects and therefore cannot be forged easily as there has been no neural network so efficient till date to reproduce the actual brain model. I have used a novel method of error analysis making use of Static and Dynamic Dimensionality Reduction for the experiment. The working environment of the project is MATLAB.
Achievements:
The project has helped me hone my skills in the realms of Authentication Systems. I have gained experience and can wield the resources to cover all the facets of an Authentication System, ranging from Dataset handling, matching scores of the outputs to presenting the work in a form of publishable work that is acceptable to academics. I believe this is the stepping stone towards my goal in developing a PhD in realms of Computer Vision and Machine Learning.
Research Paper: Accepted at SEMCCO 2013, will appear in LNCS,Springer. For a copy of paper, please contact me over email.