Pedestrian reactions to autonomous cars

Title:Pedestrian reactions to autonomous cars
Institute:University of Twente (HMI)
Place:Enschede The Netherlands
Type:Capita selecta and Research Topics
End date:not present
HMI ContactJamy Li


Many companies are working to develop autonomous cars, one goal of which is to create fully-autonomous cars that can drive themselves without the need for a person to be in the car. How will pedestrians react to fully-autonomous cars that don’t have a human driver? Pedestrians often use eye contact and other cues to communicate with human drivers, but this isn’t possible when there’s nobody in the car. Therefore, it’s interesting to see how people react to these types of cars.

In this project, the student would first do video-editing of video footage we collected of people reacting to an autonomous car on a real-life intersection. It would involve clipping a series of videos to isolate the interaction events (i.e., human reactions to the car) in the video, and blurring the faces of some of the people in the video. This part would take approximately (10-30 hours) 1-4 days. The student needs some skills and experience in video editing to do this.

For the second part of the project, the student would try to answer: how do pedestrians who are about to cross the street react to an autonomous car that approaches the intersection? The way they answer the question is by looking through the set of video clips that capture real people reacting to an autonomous car. The student would determine how to sort those reactions into groups. The student would have to look at how best to sort the clips into groups, based on the response of the person in the clip, and determine the notable features of each of those groups. Then they would label them. They could also test whether there are some interesting similar characteristics in the people within each group.

Links to related literature:

Rothenbücher, D., Li, J., Sirkin, D., Mok, B., & Ju, W. (2015, September). Ghost driver: a platform for investigating interactions between pedestrians and driverless vehicles. In Adjunct Proceedings of the 7th International Conference on Automotive User Interfaces and Interactive Vehicular Applications (pp. 44-49). ACM.

Link to paper: