Searching for optimal sensor setting for intention recognition of vehicle driving

We have been concerned with a recognition of partial and short term intentions, "i.e. micro-level intention recognition," to which conventional methods are not applicable because of action duration uncertainty. Our approach to micro-level intention recognition is to deal with a vehicle driving task as a typical application and to apply a pattern learning function which discovers and utilizes synchronous and temporal relations among the multi-dimensional time-series data of both user and environment. In this paper, appropriate sensor settings of the vehicle for intention recognition are examined by Support Vector Machine (SVM).