IUI Logo

Automotive IUI

Augmented Reality

In the research project EyeVIUS, which is part of the BMBF-funded project SiAM, we develop new applications to enhance the road traffic safety and enrich the user experience. For this purpose we combine the functions of existing assistance systems with the latest technology in the field of driver monitoring. We test and evaluate algorithms which enable the sensor fusion between the existing car sensors and additional sensors such as eye-tracker and head-tracker. The realization of the above functions through a combination of advanced technologies in the field of user-focus extraction and environmental detection and the extraction of semantic relations between these two components.

We combine our knowledge in environment reconstruction and 3D reference resolution to offer an audiovisual augmented reality experience. This way, the reality is not only augmented with visual information, in addition the user can ask about any existing detail in the environment and receives an audiovisual response to his question.


Reference Resolution in 3D

Which object is currently in the focus of the driver? Which objects lead to distraction of drivers on a particular intersection? To find the answers to these questions, we evaluate the gaze data (and head orientation data) of the driver in three-dimensional models. This analysis is carried out automatically, so that little to no interaction by the user is necessary.



Intuitive Interaction Methods

The driver’s focus-of-attention is a key factor for building novel intuitive user interactions and enhancing the current infotainment and safety applications. For this purpose, EyeVIUS extracts and merges the information about the eye gaze and head pose of the driver. In addition, it evaluates this information in a very realistic environment to gain the best possible results for analysis and interactive applications.


Environment Modeling

We model our test and evaluation environments very precisely. For this purpose, we rely on two modeling techniques: For scanning large environments (e.g. an intersection), we use a very precise laser scanner, for scanning smaller environments (e.g. a car dashboard) we use hand-held scanners. Both techniques have millimeter precision that we can rely on to make sure that our applications and analysis software deliver the best results.