An Automatic, Adaptive Help System to Support Gestural Operation of an Automotive MMI
Ralf Nieschulz, Michael Geiger, Klaus J. Bengler, Manfred K. Lang · 2001
In this study we surveyed an approach to determine an user’s need of assistance while operating an automotive human-machine interface by means of gestures. In an extensive usability study we investigated three characteristic features that seemed to be suitable for this approach. The first stage of our 2-stage methodology is a neural network based classification that determines if a user needs help while performing a certain task. In the second stage a heuristic postprocess based on statistics determines which help this user actually needs in the given context. This approach was then evaluated in an offline application. 1.