CLUST : simulating realistic crowd behaviour by mining pattern from crowd videos

In this paper, we present a data‐driven approach to simulate realistic locomotion of virtual pedestrians. We focus on simulating low‐level pedestrians' motion, where a pedestrian's motion is mainly affected by other pedestrians and static obstacles nearby, and the preferred velocities of a...

Full description

Saved in:
Bibliographic Details
Main Authors: Zhao, Mingbi, Cai, Wentong, Turner, S. J.
Other Authors: School of Computer Science and Engineering
Format: Article
Language:English
Published: 2019
Subjects:
Online Access:https://hdl.handle.net/10356/83271
http://hdl.handle.net/10220/50079
Tags: Add Tag
No Tags, Be the first to tag this record!
Institution: Nanyang Technological University
Language: English
id sg-ntu-dr.10356-83271
record_format dspace
spelling sg-ntu-dr.10356-832712020-03-07T11:48:55Z CLUST : simulating realistic crowd behaviour by mining pattern from crowd videos Zhao, Mingbi Cai, Wentong Turner, S. J. School of Computer Science and Engineering Behavioural Animation Animation Engineering::Computer science and engineering In this paper, we present a data‐driven approach to simulate realistic locomotion of virtual pedestrians. We focus on simulating low‐level pedestrians' motion, where a pedestrian's motion is mainly affected by other pedestrians and static obstacles nearby, and the preferred velocities of agents (direction and speed) are obtained from higher level path planning models. Before the simulation, collision avoidance processes (i.e. examples) are extracted from videos to describe how pedestrians avoid collisions, which are then clustered using hierarchical clustering algorithm with a novel distance function to find similar patterns of pedestrians' collision avoidance behaviours. During the simulation, at each time step, the perceived state of each agent is classified into one cluster using a neural network trained before the simulation. A sequence of velocity vectors, representing the agent's future motion, is selected among the examples corresponding to the chosen cluster. The proposed CLUST model is trained and applied to different real‐world datasets to evaluate its generality and effectiveness both qualitatively and quantitatively. The simulation results demonstrate that the proposed model can generate realistic crowd behaviours with comparable computational cost. Accepted version 2019-10-03T05:10:18Z 2019-12-06T15:18:54Z 2019-10-03T05:10:18Z 2019-12-06T15:18:54Z 2017 Journal Article Zhao, M., Cai, W., & Turner, S. J. (2018). CLUST : simulating realistic crowd behaviour by mining pattern from crowd videos. Computer Graphics Forum, 37(1), 184-201. doi:10.1111/cgf.13259 0167-7055 https://hdl.handle.net/10356/83271 http://hdl.handle.net/10220/50079 10.1111/cgf.13259 en Computer Graphics Forum This is the peer reviewed version of the following article: Zhao, M., Cai, W., & Turner, S. J. (2018). CLUST : simulating realistic crowd behaviour by mining pattern from crowd videos. Computer Graphics Forum, 37(1), 184-201, which has been published in final form at http://dx.doi.org/10.1111/cgf.13259. This article may be used for non-commercial purposes in accordance with Wiley Terms and Conditions for Use of Self-Archived Versions. 17 p. application/pdf
institution Nanyang Technological University
building NTU Library
country Singapore
collection DR-NTU
language English
topic Behavioural Animation
Animation
Engineering::Computer science and engineering
spellingShingle Behavioural Animation
Animation
Engineering::Computer science and engineering
Zhao, Mingbi
Cai, Wentong
Turner, S. J.
CLUST : simulating realistic crowd behaviour by mining pattern from crowd videos
description In this paper, we present a data‐driven approach to simulate realistic locomotion of virtual pedestrians. We focus on simulating low‐level pedestrians' motion, where a pedestrian's motion is mainly affected by other pedestrians and static obstacles nearby, and the preferred velocities of agents (direction and speed) are obtained from higher level path planning models. Before the simulation, collision avoidance processes (i.e. examples) are extracted from videos to describe how pedestrians avoid collisions, which are then clustered using hierarchical clustering algorithm with a novel distance function to find similar patterns of pedestrians' collision avoidance behaviours. During the simulation, at each time step, the perceived state of each agent is classified into one cluster using a neural network trained before the simulation. A sequence of velocity vectors, representing the agent's future motion, is selected among the examples corresponding to the chosen cluster. The proposed CLUST model is trained and applied to different real‐world datasets to evaluate its generality and effectiveness both qualitatively and quantitatively. The simulation results demonstrate that the proposed model can generate realistic crowd behaviours with comparable computational cost.
author2 School of Computer Science and Engineering
author_facet School of Computer Science and Engineering
Zhao, Mingbi
Cai, Wentong
Turner, S. J.
format Article
author Zhao, Mingbi
Cai, Wentong
Turner, S. J.
author_sort Zhao, Mingbi
title CLUST : simulating realistic crowd behaviour by mining pattern from crowd videos
title_short CLUST : simulating realistic crowd behaviour by mining pattern from crowd videos
title_full CLUST : simulating realistic crowd behaviour by mining pattern from crowd videos
title_fullStr CLUST : simulating realistic crowd behaviour by mining pattern from crowd videos
title_full_unstemmed CLUST : simulating realistic crowd behaviour by mining pattern from crowd videos
title_sort clust : simulating realistic crowd behaviour by mining pattern from crowd videos
publishDate 2019
url https://hdl.handle.net/10356/83271
http://hdl.handle.net/10220/50079
_version_ 1681037704888844288