Human gaze behavior during walking in a natural environment

Author: M. HasenjaegerAffiliation: Honda Research Institute Europe GmbHAddress: Carl-Legien-Str. 30, 63065 Offenbach am Main, GermanyCorresponding author: M. HasenjaegerContact information: [e-mail] introductionThis dataset contains data collected in everyday walk scenarios in a natural urban environment. It is a subset of the data published in [1] that has been supplemented by IMU data recorded with the mobile eye-tracker.The data is being made public as supplementary data for publications and in order for other researchers to use these data in their own work.MethodsThe data were recorded from 20 healthy participants (five females, fifteen males, between 18 and 69 years old, 178.5 ± 7.64 cm, 72.9 ± 8.7 kg).Data recordings took place around a suburban train station close to Frankfurt/Main in Germany. All participants provided written informed consent, including written permission to publish the data of the study. The study was approved by the Bioethics Committee in Honda’s R&amp;D (97HM-036H, Dec. 14, 2020).The data recording procedure is described in detail in [2]. From the three different walk courses described in [2], we only include data from course A for each of the 20 participants. The data is augmented by the data of the head pitch angle that is available from the IMU sensor of the mobile eyetracker. These data are not included with [1]. From the motion data in [1], we only include the lower body joint angles.Baseline pitch angles of zero degrees for head and eyes, respectively, were established empirically for each individual as the average pitch angles measured during straight level walking in segment 3 of course A, cf. Fig. 8 in [2]. The data from the first and last six steps in this segment were considered to be part of walk mode transitions and hence were not used for the computation of the baseline pitch angles. Deviations of the head and eye pitch angles from these empirical baselines are included in this dataset.Transitions between walk modes are labeled more exactly than in [1].Description of the data in this datasetThe data in this dataset has been organized by participant.For each participant, there is a file "gaze_motion_courseA_&lt;<i>participant ID&gt;</i>" in CSV format containing, among others, the following information:time stamp of measurementright and left foot x-, y-, z-position coordinatesright and left hip, knee, and ankle angleshead pitch angleeye pitch anglegaze pitch angleheel strike indicatorsstep indicatorswalk mode, i.e. walk, stairs up, stairs down, ramp up, ramp downindicator of change in walk modeUPDATE (version 1)In the first published version there is an error in the calculation of the baseline values which resulted in the inclusion of steps that were part of the transition phase. This error is corrected in this version (version 1). It affects only columns containing deviations from baseline values. These deviations have been corrected.Deviations from straight level walking for the lower body joint angles are included in this data set.<br>References[1] Losing, V., Hasenjäger, M.(2022). NEWBEE: A Multi-Modal Gait Database of Natural Everyday-Walk in an Urban Environment. figshare. Collection. doi: 10.6084/m9.figshare.c.5758997.v1[2] Losing, V., &amp; Hasenjäger, M. (2022). A multi-modal gait database of natural everyday-walk in an urban environment. Scientific Data, 9(1), 473. doi: 10.1038/s41597-022-01580-3<br>