Crowd Intelligence (CRIN)

April 2018 - July 2018
Interdisziplinary project of the Weißensee Kunsthochschule Berlin, Fachhochschule Potsdam and the BMW Group Design
In collaboration with Bela Kurek
Supervised by Prof. Nils Krüger, Prof. Dr. Frank Heidmann

The success of autonomous vehicles depends on the accuracy of image processing systems that permanently capture and evaluate the environment of a vehicle. Recurring accidents with autopilots show that automated driving does not work as well as expected.
A possible approach to more training data uses the collective intelligence of the crowd, which trains AI systems in the vehicle in real time and thus contributes to a permanent increase in recognition accuracy and safety.