The transference of thought pulses (electrical signals) into hardware, has accelerated the evolution of Driver Assistance Systems (DAS) in the automobile. The more 'control' gets transferred to microprocessors (Electronic Control Units - ECUs) the more autonomous the vehicle becomes, dipping driver participation. A necessary but sufficient condition for full autonomy is the ability to drive as humans do. There is the need therefore for driverless vehicles to detect and respond to exaggerated ad hoc bodily movements that humans use to avoid road accidents. In other words, cars of the near future have to inteprete gestures such as when a traffic police officer beckons a stop with a stretched arm or a traffic marshal hoists a flag and anticipate random activities that may result in road accidents. The research seeks to advance the dependability of the state-of-the-art traffic scene extrapolation and interpretation technology in identifying a gesture command source at a random traffic scene and initiate an appropriate response in the automobile. This enhances the perception of localized road users as communication partners and not as objects to avoid during autonomous driving.
All scholarship related funding is provided through the kind courtesy of The Deutscher Akademischer Austauschdienst (DAAD)