Modeling human spatial orientation and motion perception

We here present one part of a generic spatial orientation and motion sickness model. The part focussed on here describes visual-vestibular interactions regarding motion and attitude perception. The key issue regarding the processing of vestibular cues concerns the way accelerations due to motion are separated from those by gravity. Here we take a simple low-pass filter to solve this problem. We separate visual information in circular and linear flow-information, and in frame information. The circular optic flow is just added to vestibular canal signals. Linear optic flow is weighted with high-pass filtered otolith afferents. Visual frame information is weighted with vestibular and longitudinal body-axis information. To evaluate this model, we simulated a simulator take-off experiment performed by Groen et al. (2000). These data gave motion filter parameters for a run that was judged to be good, and one that was judged to be bad. We here conclude that: 1) The presented theory of motion and attitude perception is applicable in optimizing motion filters. 2) Negative acceleration perceptions should be prevented. 3) With impeded vision linear velocity perception in a hexapod setup is inadequate. 4) Attitude perception is the most sensitive factor discerning good from bad simulator trials. 5) A good simulator run is probably one with linear velocity and attitude perception that are close to real, whereas acceleration or force perception is less important. 6) The presented theory can and should be extended for further optimization and to predict simulator sickness in addition.