Research

 

Intelligent Vehicle Experimental Test-beds

Human-

Centered Driver Assistance Systems

Vision Systems for Occupant Posture

 Analysis

 

Vision Systems for Driver Analysis and Interface

 

Related Research

 

Complete Context Capture Dataset

As part of the Human-Centered Driver Assistance Systems Research, a dataset consisting of of twenty different test subjects was collected.  There is about forty minutes to an hour of driving data for each test subject.  A complete capture was taken consisting of 7 video streams from inside and outside the vehicle, audio data of the driver, GPS data, and CAN bus data (including LASER RADAR information).  Each subject was also asked to fill out a questionnaire after the drive and come back later for an interview.  Below is a sample taken from this dataset.

Visualization of data collected in the Complete Context Capture Data Archive

The images clockwise from the upper left are the forward facing rectilinear camera, roof-mounted omnidirectional camera, head-mounted subject camera, and a quad composite video of the head-mounted camera, the rear-mounted rectilinear camera, and the near-IR foot camera.

The above graph is a sample of the steering angle collected from the vehicle CAN bus data.