Print Email Facebook Twitter Relating Human Gaze and Manual Control Behavior in Preview Tracking Tasks with Spatial Occlusion Title Relating Human Gaze and Manual Control Behavior in Preview Tracking Tasks with Spatial Occlusion Author Rezunenko, Evgeny (Student TU Delft) van der El, Kasper (TU Delft Control & Simulation) Pool, D.M. (TU Delft Control & Simulation) van Paassen, M.M. (TU Delft Control & Simulation) Mulder, Max (TU Delft Control & Operations) Department Control & Operations Date 2018 Abstract In manual tracking tasks with preview of the target trajectory, humans have been modeled as dual-mode “near” and “far” viewpoint controllers. This paper investigates the physical basis of these two control mechanisms, and studies whether estimated viewpoint positions represent those parts of the previewed trajectory which humans use for control. A combination of human gaze and control data is obtained, through an experiment which compared tracking with full preview (1.5 s), occluded preview, and no preview. System identification is applied to estimate the two look-ahead time parameters of a two-viewpoint preview model. Results show that humans focus their gaze often around the model’s near-viewpoint position, and seldom at the far viewpoint. Gaze measurements may augment control data for the online identification of preview control behavior, to improve personalized monitoring or shared-control systems in vehicles. To reference this document use: http://resolver.tudelft.nl/uuid:89e20e80-af13-46bf-9525-750f634c1035 DOI https://doi.org/10.1109/SMC.2018.00583 Source Proceedings of the IEEE International Conference on Systems, Man, and Cybernetics: Myazaki, Japan, 2018 Event SMC 2018: IEEE International Conference on Systems, Man, and Cybernetics, 2018-10-07 → 2018-10-10, Myazaki, Japan Part of collection Institutional Repository Document type conference paper Rights © 2018 Evgeny Rezunenko, Kasper van der El, D.M. Pool, M.M. van Paassen, Max Mulder Files PDF 2018_Rezunenko_ProcIEEE_SMC.pdf 529.1 KB Close viewer /islandora/object/uuid:89e20e80-af13-46bf-9525-750f634c1035/datastream/OBJ/view