Efficient trajectory extraction and parameter learning for data-driven crowd simulation

We present a trajectory extraction and behavior-learning algorithm for data-driven crowd simulation. Our formulation is based on incrementally learning pedestrian motion models and behaviors from crowd videos. We combine this learned crowd-simulation model with an online tracker based on particle filtering to compute accurate, smooth pedestrian trajectories. We refine this motion model using an optimization technique to estimate the agents' simulation parameters. We highlight the benefits of our approach for improved data-driven crowd simulation, including crowd replication from videos and merging the behavior of pedestrians from multiple videos. We highlight our algorithm's performance in various test scenarios containing tens of human-like agents.

[1]  H. Zha,et al.  A fully online and unsupervised system for large and high-density area surveillance: Tracking, semantic scene learning and abnormality detection , 2013, TIST.

[2]  Dinesh Manocha,et al.  REACH - Realtime crowd tracking using a hybrid motion model , 2015, 2015 IEEE International Conference on Robotics and Automation (ICRA).

[3]  Stéphane Donikian,et al.  Online inserting virtual characters into dynamic video scenes , 2011, Comput. Animat. Virtual Worlds.

[4]  Dariu Gavrila,et al.  Monocular Pedestrian Detection: Survey and Experiments , 2009, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[5]  Dinesh Manocha,et al.  AdaPT: Real-time adaptive pedestrian tracking for crowded scenes , 2014, 2014 IEEE International Conference on Robotics and Automation (ICRA).

[6]  Norman I. Badler,et al.  Efficient motion retrieval in large motion databases , 2013, I3D '13.

[7]  Gérard G. Medioni,et al.  Tracking Using Motion Patterns for Very Crowded Scenes , 2012, ECCV.

[8]  Dinesh Manocha,et al.  Reciprocal n-Body Collision Avoidance , 2011, ISRR.

[9]  Xiaona Li,et al.  Simulating realistic crowd based on agent trajectories , 2013, Comput. Animat. Virtual Worlds.

[10]  Dinesh Manocha,et al.  Leveraging Long-Term Predictions and Online Learning in Agent-Based Multiple Person Tracking , 2014, IEEE Transactions on Circuits and Systems for Video Technology.

[11]  Dinesh Manocha,et al.  Realtime Multilevel Crowd Tracking Using Reciprocal Velocity Obstacles , 2014, 2014 22nd International Conference on Pattern Recognition.

[12]  Rainer Stiefelhagen,et al.  Evaluating Multiple Object Tracking Performance: The CLEAR MOT Metrics , 2008, EURASIP J. Image Video Process..

[13]  Lei Zhang,et al.  Real-Time Compressive Tracking , 2012, ECCV.

[14]  Dinesh Manocha,et al.  Directing Crowd Simulations Using Navigation Fields , 2011, IEEE Transactions on Visualization and Computer Graphics.

[15]  Daniel Cohen-Or,et al.  Data Driven Evaluation of Crowds , 2009, MIG.

[16]  Dinesh Manocha,et al.  Real-time Crowd Tracking using Parameter Optimized Mixture of Motion Models , 2014, ArXiv.

[17]  Ko Nishino,et al.  Going with the Flow: Pedestrian Efficiency in Crowded Scenes , 2012, ECCV.

[18]  P. Torrens,et al.  Building Agent‐Based Walking Models by Machine‐Learning on Diverse Databases of Space‐Time Trajectory Samples , 2011 .

[19]  Jehee Lee,et al.  Morphable crowds , 2010, SIGGRAPH 2010.

[20]  Soraia Raupp Musse,et al.  Using computer vision to simulate the motion of virtual agents , 2007, Comput. Animat. Virtual Worlds.

[21]  A. Tyagi,et al.  A Context-Based Tracker Switching Framework , 2008, 2008 IEEE Workshop on Motion and video Computing.

[22]  Yi Li,et al.  Cloning crowd motions , 2012, SCA '12.

[23]  Petros Faloutsos,et al.  SteerFit: automated parameter fitting for steering algorithms , 2014, SCA '14.

[24]  Pramod Sharma,et al.  Unsupervised incremental learning for improved object detection in a video , 2012, 2012 IEEE Conference on Computer Vision and Pattern Recognition.

[25]  Yi Wu,et al.  Online Object Tracking: A Benchmark , 2013, 2013 IEEE Conference on Computer Vision and Pattern Recognition.

[26]  Dinesh Manocha,et al.  A statistical similarity measure for aggregate crowd dynamics , 2012, ACM Trans. Graph..

[27]  Ivan Laptev,et al.  Density-aware person detection and tracking in crowds , 2011, ICCV.

[28]  Qunsheng Peng,et al.  Inserting virtual pedestrians into pedestrian groups video with behavior consistency , 2013, The Visual Computer.

[29]  Mubarak Shah,et al.  Floor Fields for Tracking in High Density Crowd Scenes , 2008, ECCV.

[30]  Neil J. Gordon,et al.  A tutorial on particle filters for online nonlinear/non-Gaussian Bayesian tracking , 2002, IEEE Trans. Signal Process..

[31]  Dinesh Manocha,et al.  Parameter estimation and comparative evaluation of crowd simulations , 2014, Comput. Graph. Forum.

[32]  Dimitris N. Metaxas,et al.  Eurographics/ Acm Siggraph Symposium on Computer Animation (2007) Group Behavior from Video: a Data-driven Approach to Crowd Simulation , 2022 .

[33]  Ivan Laptev,et al.  Data-driven crowd analysis in videos , 2011, ICCV.

[34]  Dani Lischinski,et al.  Crowds by Example , 2007, Comput. Graph. Forum.