A multi-sensor dataset of human-human handover

摘要: Abstract The article describes a multi-sensor dataset of human-human handovers composed over 1000 recordings collected from 18 volunteers. refer to 76 test configurations, which consider different volunteer׳s starting positions and roles, objects pass motion strategies. In all experiments, we acquire 6-axis inertial data two smartwatches, the 15-joint skeleton model one volunteer with an RGB-D camera upper-body both persons using total 20 capture markers. are annotated videos questionnaires about perceived characteristics handover.