In this master thesis, we present a new framework for the automatic detection, recognition and learning of human faces with random forests. The principal motivation for our work is towards the design of a fast face processing system, especially applicable to mobile robots. Given a high amount of social interaction between a robot and human beings, the experience is rendered more natural if the robot is able to detect faces and recognize corresponding identities. If the identity of the person is not known then it is learnt by the robot.
|Publication status||Published - 2008|