A multi-sensor dataset of human-human handover

Published: 20-04-2020| Version 3 | DOI: 10.17632/cyr9gmhspy.3
Alessandro Carfì,
Francesco Foglino,
Barbara Bruno,
Fulvio Mastrogiovanni


The dataset contains over 1000 recordings of human-human handovers collected from 18 volunteers. The recordings refer to 76 test configurations, which consider volunteer's starting position and role, object to pass and experimenter's motion strategy. Data collected come from 6-axis IMU one for each volunteer, Microsoft Kinect for skeleton tracking and a motion capture system. Each recording is also annotated with a video and a questionnaire about the perceived characteristics of the handover.


Steps to reproduce

For further details about the dataset collection procedure, refer to the article "A multi-sensor dataset of human-human handover" published on Data in Brief (https://doi.org/10.1016/j.dib.2018.11.110).