Driver Intent Research
Driver intent estimation is important for a variety
of aspects of intelligent vehicles. It can used in lane departure
systems and collision avoidance systems as well as many other types of
applications. Part of determining driver intent is determining
when the driver is about to make a lane change. This is not always
as easy as detecting when the driver has turned on his turn signal.
Below is a diagram of a typical lane change.
Using a data-driven approach through a Sparse
Bayesian Learning algorithm the following results were generated in the
form of a receiver-operator characteristics curve.