Using Facial Feature Recognition and Head Tracking to Control Games

Abstract

Many interfaces exist to control computer games, both for desktop computers and for consoles. For desktop gaming, the most commonly used interface is the mouse and keyboard. Most alternative interfaces often require hardware that users do not already possess, such as gamepads, joysticks, a Kinect, or a Virtual Reality-device. If games want to use alternative interfaces they often have to be specifically designed to be able to do so. In this work is presented a novel interface `FaceDriver' to control games. FaceDriver makes use of head tracking, and only requires a webcam. This interface is usable on a large number of existing games, as it allows games to interpret the interface as if it were an interface for which the game was designed. A user study is conducted in which 25 users play a racing game using this interface. The in-game performance of this interface is compared to traditional interfaces, and the user engagement is evaluated. The results show, that while the interface isn't competitive in terms of in-game performance compared to a mouse and keyboard, users do prefer FaceDriver over traditional interfaces.