The Role of Data-driven Priors in Multi-agent Crowd Trajectory Estimation

Trajectory interpolation, the process of filling-in the gaps and removing noise from observed agent trajectories, is an essential task for the motion inference in multi-agent setting. A desired trajectory interpolation method should be robust to noise, changes in environments or agent densities, while also being yielding realistic group movement behaviors. Such realistic behaviors are, however, challenging to model as they require avoidance of agent-agent or agent-environment collisions and, at the same time, seek computational efficiency. In this paper, we propose a novel framework composed of data-driven priors (local, global or combined) and an efficient optimization strategy for multi-agent trajectory interpolation. The data-driven priors implicitly encode the dependencies of movements of multiple agents and the collision-avoiding desiderata, enabling elimination of costly pairwise collision constraints and resulting in reduced computational complexity and often improved estimation. Various combinations of priors and optimization algorithms are evaluated in comprehensive simulated experiments. Our experimental results reveal important insights, including the significance of the global flow prior and the lesser-than-expected influence of data-driven collision priors.

[1]  Bingbing Ni,et al.  Crowded Scene Analysis: A Survey , 2015, IEEE Transactions on Circuits and Systems for Video Technology.

[2]  Luc Van Gool,et al.  Tracking with a mixed continuous-discrete Conditional Random Field , 2013, Comput. Vis. Image Underst..

[3]  Van GoolLuc,et al.  Tracking with a mixed continuous-discrete Conditional Random Field , 2013 .

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

[5]  Javier Alonso-Mora,et al.  A message-passing algorithm for multi-agent trajectory planning , 2013, NIPS.

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

[7]  Glenn Reinman,et al.  An Open Framework for Developing, Evaluating, and Sharing Steering Algorithms , 2009, MIG.

[8]  Glenn Reinman,et al.  SteerBench: a benchmark suite for evaluating steering behaviors , 2009 .

[9]  Dinesh Manocha,et al.  Modeling, Simulation and Visual Analysis of Crowds: A Multidisciplinary Perspective , 2013, Modeling, Simulation and Visual Analysis of Crowds.

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

[11]  Sergio A. Velastin,et al.  Crowd analysis: a survey , 2008, Machine Vision and Applications.

[12]  Soraia Raupp Musse,et al.  Crowd Analysis Using Computer Vision Techniques , 2010, IEEE Signal Processing Magazine.

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

[14]  Takeo Kanade,et al.  Tracking in unstructured crowded scenes , 2009, 2009 IEEE 12th International Conference on Computer Vision.

[15]  Luc Van Gool,et al.  Wrong turn - No dead end: A stochastic pedestrian motion model , 2010, 2010 IEEE Computer Society Conference on Computer Vision and Pattern Recognition - Workshops.

[16]  Lee SmithBattle,et al.  A Wrong Turn , 2000 .

[17]  C. Striebel,et al.  On the maximum likelihood estimates for linear dynamic systems , 1965 .

[18]  Dinesh Manocha,et al.  Reciprocal Velocity Obstacles for real-time multi-agent navigation , 2008, 2008 IEEE International Conference on Robotics and Automation.

[19]  Irfan A. Essa,et al.  Gaussian process regression flow for analysis of motion trajectories , 2011, 2011 International Conference on Computer Vision.

[20]  Norman I. Badler,et al.  Virtual Crowds: Steps Toward Behavioral Realism , 2015, Virtual Crowds: Steps Toward Behavioral Realism.

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

[22]  Luc Van Gool,et al.  You'll never walk alone: Modeling social behavior for multi-target tracking , 2009, 2009 IEEE 12th International Conference on Computer Vision.

[23]  Vladimir Pavlovic,et al.  Filling in the blanks: reconstructing microscopic crowd motion from multiple disparate noisy sensors , 2016, 2016 IEEE Winter Applications of Computer Vision Workshops (WACVW).

[24]  Dinesh Manocha,et al.  Online parameter learning for data-driven crowd simulation and content generation , 2016, Comput. Graph..

[25]  Fei-Fei Li,et al.  Socially-Aware Large-Scale Crowd Forecasting , 2014, 2014 IEEE Conference on Computer Vision and Pattern Recognition.

[26]  Helbing,et al.  Social force model for pedestrian dynamics. , 1995, Physical review. E, Statistical physics, plasmas, fluids, and related interdisciplinary topics.

[27]  Mubarak Shah,et al.  Learning motion patterns in crowded scenes using motion flow field , 2008, 2008 19th International Conference on Pattern Recognition.

[28]  Dinesh Manocha,et al.  Efficient trajectory extraction and parameter learning for data-driven crowd simulation , 2015, Graphics Interface.

[29]  Glenn Reinman,et al.  SteerBench: a benchmark suite for evaluating steering behaviors , 2009, Comput. Animat. Virtual Worlds.

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

[31]  Jia Pan,et al.  Deep-Learned Collision Avoidance Policy for Distributed Multiagent Navigation , 2016, IEEE Robotics and Automation Letters.