Driver distraction detection with Deep Learning

  • Sponsored: BMW AG
  • Scientific Lead: Dr. Benny Kneissl, Dominik Schniertshauer, Konrad Dongus and Joshua Görner.
  • Project Lead: Dr. Ricardo Acevedo Cabra
  • Results: The results of this project were explained in detail in the final documentation.

One of the major factors in loss of life, in particular on highways, is driver distraction. Although the automotive industry is already on its way to make self-driving cars reality, until and including Autonomous Driving level 3 the driver will still be the back-up in potentially unclear situations. The question arises how long it will take to take over the control of the vehicle in different distraction modes. In this project, you will analyze video data to detect distraction (or even the level of distraction) of the driver with the help of Deep Learning. Identifying situations of smartphone usage, 'hands-on-wheel' or gaze detection are only a few possible distraction scenarios you could focus on.