Automatically detecting the small group structure of a crowd

Research output: Chapter in Book/Report/Conference proceedingConference contribution

64 Citations (Scopus)

Abstract

Recent work on computer vision analysis of crowds tends to focus on robustly tracking individuals through the crowd or on analyzing the overall pattern of flow. Our work seeks a deeper analysis of social behavior by identifying the small group structure of crowds, forming the basis for mid-level activity analysis at the granularity of human social groups. Building upon state-of-the-art algorithms for pedestrian detection and multi-object tracking, and inspired by social science models of human collective behavior, we automatically detect small groups of individuals who are traveling together. These groups are discovered using a bottom-up hierarchical clustering approach that compares sets of individuals based on a generalized, symmetric Hausdorff distance defined with respect to pairwise proximity and velocity. We validate our results quantitatively and qualitatively on videos of real-world pedestrian scenes. Where human-coded ground truth is available, we find substantial statistical agreement between our results and the human-perceived small group structure of the crowd.

Original languageEnglish (US)
Title of host publication2009 Workshop on Applications of Computer Vision, WACV 2009
DOIs
StatePublished - Dec 1 2009
Event2009 Workshop on Applications of Computer Vision, WACV 2009 - Snowbird, UT, United States
Duration: Dec 7 2009Dec 8 2009

Publication series

Name2009 Workshop on Applications of Computer Vision, WACV 2009

Other

Other2009 Workshop on Applications of Computer Vision, WACV 2009
CountryUnited States
CitySnowbird, UT
Period12/7/0912/8/09

Fingerprint

Social sciences
Computer vision

All Science Journal Classification (ASJC) codes

  • Computer Science Applications
  • Computer Vision and Pattern Recognition

Cite this

Ge, W., Collins, R. T., & Ruback, B. (2009). Automatically detecting the small group structure of a crowd. In 2009 Workshop on Applications of Computer Vision, WACV 2009 [5403123] (2009 Workshop on Applications of Computer Vision, WACV 2009). https://doi.org/10.1109/WACV.2009.5403123
Ge, Weina ; Collins, Robert T. ; Ruback, Barry. / Automatically detecting the small group structure of a crowd. 2009 Workshop on Applications of Computer Vision, WACV 2009. 2009. (2009 Workshop on Applications of Computer Vision, WACV 2009).
@inproceedings{e36ea38b2bbf40309d1be37fd959372c,
title = "Automatically detecting the small group structure of a crowd",
abstract = "Recent work on computer vision analysis of crowds tends to focus on robustly tracking individuals through the crowd or on analyzing the overall pattern of flow. Our work seeks a deeper analysis of social behavior by identifying the small group structure of crowds, forming the basis for mid-level activity analysis at the granularity of human social groups. Building upon state-of-the-art algorithms for pedestrian detection and multi-object tracking, and inspired by social science models of human collective behavior, we automatically detect small groups of individuals who are traveling together. These groups are discovered using a bottom-up hierarchical clustering approach that compares sets of individuals based on a generalized, symmetric Hausdorff distance defined with respect to pairwise proximity and velocity. We validate our results quantitatively and qualitatively on videos of real-world pedestrian scenes. Where human-coded ground truth is available, we find substantial statistical agreement between our results and the human-perceived small group structure of the crowd.",
author = "Weina Ge and Collins, {Robert T.} and Barry Ruback",
year = "2009",
month = "12",
day = "1",
doi = "10.1109/WACV.2009.5403123",
language = "English (US)",
isbn = "9781424454976",
series = "2009 Workshop on Applications of Computer Vision, WACV 2009",
booktitle = "2009 Workshop on Applications of Computer Vision, WACV 2009",

}

Ge, W, Collins, RT & Ruback, B 2009, Automatically detecting the small group structure of a crowd. in 2009 Workshop on Applications of Computer Vision, WACV 2009., 5403123, 2009 Workshop on Applications of Computer Vision, WACV 2009, 2009 Workshop on Applications of Computer Vision, WACV 2009, Snowbird, UT, United States, 12/7/09. https://doi.org/10.1109/WACV.2009.5403123

Automatically detecting the small group structure of a crowd. / Ge, Weina; Collins, Robert T.; Ruback, Barry.

2009 Workshop on Applications of Computer Vision, WACV 2009. 2009. 5403123 (2009 Workshop on Applications of Computer Vision, WACV 2009).

Research output: Chapter in Book/Report/Conference proceedingConference contribution

TY - GEN

T1 - Automatically detecting the small group structure of a crowd

AU - Ge, Weina

AU - Collins, Robert T.

AU - Ruback, Barry

PY - 2009/12/1

Y1 - 2009/12/1

N2 - Recent work on computer vision analysis of crowds tends to focus on robustly tracking individuals through the crowd or on analyzing the overall pattern of flow. Our work seeks a deeper analysis of social behavior by identifying the small group structure of crowds, forming the basis for mid-level activity analysis at the granularity of human social groups. Building upon state-of-the-art algorithms for pedestrian detection and multi-object tracking, and inspired by social science models of human collective behavior, we automatically detect small groups of individuals who are traveling together. These groups are discovered using a bottom-up hierarchical clustering approach that compares sets of individuals based on a generalized, symmetric Hausdorff distance defined with respect to pairwise proximity and velocity. We validate our results quantitatively and qualitatively on videos of real-world pedestrian scenes. Where human-coded ground truth is available, we find substantial statistical agreement between our results and the human-perceived small group structure of the crowd.

AB - Recent work on computer vision analysis of crowds tends to focus on robustly tracking individuals through the crowd or on analyzing the overall pattern of flow. Our work seeks a deeper analysis of social behavior by identifying the small group structure of crowds, forming the basis for mid-level activity analysis at the granularity of human social groups. Building upon state-of-the-art algorithms for pedestrian detection and multi-object tracking, and inspired by social science models of human collective behavior, we automatically detect small groups of individuals who are traveling together. These groups are discovered using a bottom-up hierarchical clustering approach that compares sets of individuals based on a generalized, symmetric Hausdorff distance defined with respect to pairwise proximity and velocity. We validate our results quantitatively and qualitatively on videos of real-world pedestrian scenes. Where human-coded ground truth is available, we find substantial statistical agreement between our results and the human-perceived small group structure of the crowd.

UR - http://www.scopus.com/inward/record.url?scp=77951153793&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=77951153793&partnerID=8YFLogxK

U2 - 10.1109/WACV.2009.5403123

DO - 10.1109/WACV.2009.5403123

M3 - Conference contribution

AN - SCOPUS:77951153793

SN - 9781424454976

T3 - 2009 Workshop on Applications of Computer Vision, WACV 2009

BT - 2009 Workshop on Applications of Computer Vision, WACV 2009

ER -

Ge W, Collins RT, Ruback B. Automatically detecting the small group structure of a crowd. In 2009 Workshop on Applications of Computer Vision, WACV 2009. 2009. 5403123. (2009 Workshop on Applications of Computer Vision, WACV 2009). https://doi.org/10.1109/WACV.2009.5403123