Drowsy Driving Dataset

Department of Computer Science
University of North Carolina at Chapel Hill
Chapel Hill, NC 27599-3175

This dataset is part of the multi-institution project VeHICaL: Verified Human Interfaces, Control, and Learning for Semi-Autonomous Systems.
This dataset is owned and managed by Alyssa Byrnes and Dr. Cynthia Sturton.

To access this dataset, please fill out this form.
If you haven't gotten access after a few days, please email abyrnes1@cs.unc.edu.
photo of experimental setup


First, subjects filled out a survey to indicate their current and average drowsiness levels. The questions on this survey were determined using drowsiness information that coincided with traffic accidents. Then the subjects' faces were filmed while they perform the following activities: reading a passage, watching a video of a driving simulation, and performing a driving simulation.

The driving simulation involved users driving around in ``freeplay mode'' in Driving Simulator 2013 using Logitech Driving Force G29 Racing Wheel for PlayStation 4 and PlayStation 3.

This work is part of the paper Using a Driver's Eye Data to Predict Accident-Causing Drowsiness Levels; Alyssa Byrnes and Cynthia Sturton; The 21st IEEE International Conference on Intelligent Transportation Systems, November 2018.