Scalable Detection of Crowd Motion Patterns

Research output: Contribution to journalArticleAcademicpeer-review

1 Citation (Scopus)
7 Downloads (Pure)

Abstract

Studying the movements of crowds is important for understanding and predicting the behavior of large groups of people. When analyzing such crowds, one is often interested in the long-term macro-level motions of the crowd, as opposed to the micro-level individual movements at each moment in time. A high-level representation of these motions is thus desirable. In this work, we present a scalable method for detection of crowd motion patterns, i.e., spatial areas describing the dominant motions within the crowd. For measuring crowd movements, we propose a fast, scalable, and low-cost method based on proximity graphs. For analyzing crowd movements, we utilize a three-stage pipeline: (1) represents the behavior of each person at each moment in time using a low-dimensional data point, (2) cluster these data points based on spatial relations, and (3) concatenate these clusters based on temporal relations. Experiments on synthetic datasets reveals our method can handle various scenarios including curved lanes and diverging flows. Evaluation on real-world datasets shows our method is able to extract useful motion patterns from such scenarios which could not be properly detected by existing methods. Overall, we see our work as an initial step towards rich pattern recognition.
Original languageEnglish
JournalIEEE transactions on knowledge and data engineering
DOIs
Publication statusE-pub ahead of print/First online - 1 Nov 2018

Fingerprint

Pattern recognition
Macros
Pipelines
Costs
Experiments

Keywords

  • Crowd Analytics
  • Pedestrian Dynamics
  • Spatio-Temporal Clustering
  • Trajectory Clustering
  • Proximity Graph

Cite this

@article{d7b8ca046d354860a8bca7d6a0bb5143,
title = "Scalable Detection of Crowd Motion Patterns",
abstract = "Studying the movements of crowds is important for understanding and predicting the behavior of large groups of people. When analyzing such crowds, one is often interested in the long-term macro-level motions of the crowd, as opposed to the micro-level individual movements at each moment in time. A high-level representation of these motions is thus desirable. In this work, we present a scalable method for detection of crowd motion patterns, i.e., spatial areas describing the dominant motions within the crowd. For measuring crowd movements, we propose a fast, scalable, and low-cost method based on proximity graphs. For analyzing crowd movements, we utilize a three-stage pipeline: (1) represents the behavior of each person at each moment in time using a low-dimensional data point, (2) cluster these data points based on spatial relations, and (3) concatenate these clusters based on temporal relations. Experiments on synthetic datasets reveals our method can handle various scenarios including curved lanes and diverging flows. Evaluation on real-world datasets shows our method is able to extract useful motion patterns from such scenarios which could not be properly detected by existing methods. Overall, we see our work as an initial step towards rich pattern recognition.",
keywords = "Crowd Analytics, Pedestrian Dynamics, Spatio-Temporal Clustering, Trajectory Clustering, Proximity Graph",
author = "Heldens, {Stijn Jespah} and Litvak, {Nelli Vladimirovna} and {van Steen}, {Martinus Richardus}",
year = "2018",
month = "11",
day = "1",
doi = "10.1109/TKDE.2018.2879079",
language = "English",
journal = "IEEE transactions on knowledge and data engineering",
issn = "1041-4347",
publisher = "IEEE Computer Society",

}

Scalable Detection of Crowd Motion Patterns. / Heldens, Stijn Jespah; Litvak, Nelli Vladimirovna; van Steen, Martinus Richardus.

In: IEEE transactions on knowledge and data engineering, 01.11.2018.

Research output: Contribution to journalArticleAcademicpeer-review

TY - JOUR

T1 - Scalable Detection of Crowd Motion Patterns

AU - Heldens, Stijn Jespah

AU - Litvak, Nelli Vladimirovna

AU - van Steen, Martinus Richardus

PY - 2018/11/1

Y1 - 2018/11/1

N2 - Studying the movements of crowds is important for understanding and predicting the behavior of large groups of people. When analyzing such crowds, one is often interested in the long-term macro-level motions of the crowd, as opposed to the micro-level individual movements at each moment in time. A high-level representation of these motions is thus desirable. In this work, we present a scalable method for detection of crowd motion patterns, i.e., spatial areas describing the dominant motions within the crowd. For measuring crowd movements, we propose a fast, scalable, and low-cost method based on proximity graphs. For analyzing crowd movements, we utilize a three-stage pipeline: (1) represents the behavior of each person at each moment in time using a low-dimensional data point, (2) cluster these data points based on spatial relations, and (3) concatenate these clusters based on temporal relations. Experiments on synthetic datasets reveals our method can handle various scenarios including curved lanes and diverging flows. Evaluation on real-world datasets shows our method is able to extract useful motion patterns from such scenarios which could not be properly detected by existing methods. Overall, we see our work as an initial step towards rich pattern recognition.

AB - Studying the movements of crowds is important for understanding and predicting the behavior of large groups of people. When analyzing such crowds, one is often interested in the long-term macro-level motions of the crowd, as opposed to the micro-level individual movements at each moment in time. A high-level representation of these motions is thus desirable. In this work, we present a scalable method for detection of crowd motion patterns, i.e., spatial areas describing the dominant motions within the crowd. For measuring crowd movements, we propose a fast, scalable, and low-cost method based on proximity graphs. For analyzing crowd movements, we utilize a three-stage pipeline: (1) represents the behavior of each person at each moment in time using a low-dimensional data point, (2) cluster these data points based on spatial relations, and (3) concatenate these clusters based on temporal relations. Experiments on synthetic datasets reveals our method can handle various scenarios including curved lanes and diverging flows. Evaluation on real-world datasets shows our method is able to extract useful motion patterns from such scenarios which could not be properly detected by existing methods. Overall, we see our work as an initial step towards rich pattern recognition.

KW - Crowd Analytics

KW - Pedestrian Dynamics

KW - Spatio-Temporal Clustering

KW - Trajectory Clustering

KW - Proximity Graph

U2 - 10.1109/TKDE.2018.2879079

DO - 10.1109/TKDE.2018.2879079

M3 - Article

JO - IEEE transactions on knowledge and data engineering

JF - IEEE transactions on knowledge and data engineering

SN - 1041-4347

ER -