Dataset for driving behaviors study in the mixed traffic environment with autonomous vehicles and human-driven vehicles

Background

In the VR-enabled driving simulator environment, this dataset is collected from 50 drivers from Beihang university in China, who have different demographic information (age, gender, driving years) and prior driving-simulator and VR experience for usage. And in our study about to be submitted to a journal, it is used to study the difference of behaviors of drivers towards autonomous vehicle(AV) and human-driven vehicle(HV) in the mixed traffic flows. 

Data collection environment: CARLA + DReyeVR(https://github.com/HARPLab/DReyeVR, which offers the support for VR) + SUMO

Experiment scenario: a non-signalized crossing in the built-in map of Town 03 of CARLA

Dataset structure

There are two zip files in this dataset. The one zip file  "All_Frames_of_All_drivers.zip" is consist of each frame picture for every driver/participant during the experiment, which shows the real-time scenario in the first-person perspective of driver/participant. And the other zip file "All_excelData_of_All_drivers.zip" contains all data in excel format of all drivers/participants, which is corresponding to that in "All_Frames_of_All_drivers.zip".

The names of all files in these two zip files obey the following rules: name of driver/participant + date of his/her experiment + type of vehicle(AV or HV) + experiment times(1, 2 or 3). Taking "cmy_AV_1" for an example, it means it is "cmy" who finished the first AV experiment.