**7. Conclusions**

Utilizing eye tracking data to support sensor data fusion of MARG-sensor shows improvements of the heading accuracy in magnetically disturbed environments or for IMU sensors that do not feature a heading reference in the order of 50%. Because of the physiological coupling between eye and head rotations, eye tracking can deliver an indicator signal for near zero head orientation change. Furthermore, this trigger signal allows for individual and adaptive noise parameterization through gyroscope capturing and could be used in the context of adaptive noise estimation with respect to head motion while a sufficient amount of data is captured. The proposed method can be used with any mobile eye tracking devices that either feature a build-in IMU or MARG-sensor or are expanded by a custom or third-party sensor. The presented approach does not need a world camera and is therefore mostly independent of surrounding environmental lightning conditions. In addition, the proposed use of interchangeable north direction vector substitutes enables switching between full MARG and IMU-mode, without the need for an additional set of equations in a given filter. This guarantees a continuous quaternion computation and convergence of the filter.
