Experimental Framework for Investigating First Person Guidance and Perception

Perceptual mechanisms humans use in motion planning and guidance are responsible for their agile and versatile capabilities. Their understandings would help advance both autonomous control and improve human-machine interfaces. This paper extends prior work by collecting human trajectory and gaze data while performing simulated first-person motion guidance tasks. This data is considered under the hypothesis that satisficing concepts play a significant role in human performance and adaptability. The resulting trajectory data exhibits a hierarchical partitioning structure (interaction patterns) as shown in prior work. Invariance between sub goal partitions reveal the reuse of prior knowledge in both motion primitives and perceptual behavior. By considering vehicle state and gaze as a combined state, analysis suggests that this hierarchical structure extends to the combined action-perception process.