Abstract:For walking pedestrians, when they are blocked by obstacles or other pedestrians, they adjust their speeds and directions to avoid colliding with them, which is called collision avoidance behavior. This behavior is the most complex part of pedestrians' walking processes and its modeling and simulation are the keys to realistic crowd simulation, which serves as the foundation for various applications. However, most existing methods either lack the representation power to accurately model the complex collision behavior or do not model it explicitly, which leads to a poor level of realism of the simulation. To realize realistic crowd simulation, we propose to analyze, understand, and model the collision avoidance behavior in a data-driven way. First, to automatically detect collision avoidance behavior for further analysis, we propose a domain transformation algorithm that detects it by transforming the trajectories in the spatial domain into a new domain where the behavior is much more apparent and is thus easier to detect. The new domain also provides a new perspective for understanding collision avoidance behavior. Second, since there are no mature metrics to evaluate the level of realism, we propose a new evaluation metric based on the least-effort theory, which evaluates the realism of collision avoidance behavior by its physical and mental consumption. This evaluation metric also provides the foundation of modeling. Third, for realistic crowd simulation, we design a reinforcement learning model. It trains agents with our proposed reward function that models pedestrians' intrinsic needs of "reducing effort consumption'' and thus can guide agents to behave realistically when avoiding collisions. Extensive experiments show our model is 55.9% and 52.5% more realistic in collision avoidance behavior than the best baselines on two real-world datasets. We release our codes at https://github.com/tsinghua-fib-lab/TECRL.

A Crowd Behavior Model Based on Reciprocal Velocity Obstacle Algorithm

A cooperative mobility model for multiple autonomous vehicles

Modifications to Social Force Model

Pedestrians’ behavior in emergency evacuation: Modeling and simulation*

Visually-Guided Pedestrian Crowd Simulation

A Mixed Crowd Movement Model Incorporating Chasing Behavior

A collective motion model based on two-layer relationship mechanism for bi-direction pedestrian flow simulation

Relationship Between Behavior Aggressiveness and Pedestrian Dynamics Using Behavior-Based Cellular Automata Model.

Group Modeling: A Unified Velocity-Based Approach.

A Local Behavior Model for Small Pedestrian Groups

Behavior-based Cellular Automaton Model for Pedestrian Dynamics

An extended queueing model based on vision and morality for crowd evacuation

Crowd Dynamics of Self-Propelled Individuals with Collision Avoidance Considering Anticipation and Intrusion Aversion

A modified social force model for crowd evacuation considering collision predicting behaviors

Data-Driven and Collision-Free Hybrid Crowd Simulation Model for Real Scenario

Understanding and Modeling Collision Avoidance Behavior for Realistic Crowd Simulation

A Modified Social Force Model Considering Relative Velocity of Pedestrians

Crowd evacuation simulation based on hierarchical agent model and physics‐based character control

A Simple and Realistic Pedestrian Model for Crowd Simulation and Application

Model-based Crowd Behaviours in Human-solution Space

Crowd Dynamics Based on a Modified Social Force Model Considering Predicting Collision Behaviors