Skip to main content

Bodian Seminar: Paul R. MacNeilage, PhD

February 24, 2020 @ 4:00 pm – 5:00 pm

Bodian Seminar: Paul R. MacNeilage PhD


Department of Psychology & Institute for Neuroscience

University of Nevada, Reno

Perception of visual-vestibular conflict and characterization of natural head-eye movements

To reconstruct how the head is moving relative to the environment, the nervous system relies on a combination of visual and vestibular sensory information. Vestibular signals are driven by head movement whereas visual motion signals are driven by both head and eye movement. Knowledge of eye movement based on motor efference is therefore necessary to allow for comparison and integration of visual and vestibular cues. Motor efference signals associated with head-on-body movement can also supplement sensory estimates of head movement.

In this talk I will present results of psychophysical studies investigating how these signals interact during perception of a stable visual environment. During head movement, we manipulate the gain of environmental visual motion displayed in a virtual reality headset to measure the range of gains that is compatible with perception of a stable environment, originally dubbed the Range of Immobility by Hans Wallach. Performance is compared across active and passive head movement and across scene-fixed and head-fixed fixation conditions. We find that sensitivity to conflict is best (i.e. the range of immobility is smallest) under more natural conditions, in other words, when head movements are actively generated and when observers fixate world-fixed targets. We also observe a trade-off such that optimal cue integration comes at the cost of impaired conflict detection. To relate experimental results to natural behavior we have developed a system to track 6DOF positional head movement and binocular eye movement during everyday behavior outside the lab, and we have begun modeling and characterizing the statistics of natural visual and vestibular stimulation and motor behavior. In a Bayesian framework, these statistical measures approximate priors that can help explain known biases in perception of heading and verticality.

Watch Webcast