Data from: Prediction of Pedestrian Speed with Artificial Neural Networks

posted on 07.01.2020 by Antoine Tordeux, Mohcine Chraibi, Armin Seyfried, Andreas Schadschneider

Corridor data are trajectories of pedestrians in a closed corridor of lenght 30m and width 1.8m. The trajectories are measured on a section of length 6m. Experiments are carried out with N=15, 30, 60, 85, 95, 110, 140 and 230 participants.

Bottleneck data are trajectories of pedestrian in a bottleneck of lenght 8m and width 1.8m. Experiments are carried out with 150 participants for bottleneck widths w=0.7, 0.95 1.2 and 1.8m.

See http://ped.fz-juelich.de/experiments/2009.05.12_Duesseldorf_Messe_Hermes/docu/VersuchsdokumentationHERMES.pdf page 20 and 24 for details (in German). The data are part of the online database http://ped.fz-juelich.de/database.

Column names of the file are: ID FRAME X Y Z.

  • ID is the pedestrian ID.
  • FRAME is the frame number (frame rate is 1/16s).
  • X Y and Z are pedestrian position in 3D.