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.

Collision Avoidance Mechanism for Pedestrian Interactions

Forecasting the Evolution of Urban Mobility: the Influence of Anthropomorphism and Social Responsiveness in the Transition from Human to Automated Driving

Experimental study on pedestrians' collision avoidance

Pedestrians’ behavior in emergency evacuation: Modeling and simulation*

A modified social force model considering collision avoidance based on empirical studies

Safety Envelope of Pedestrians Upon Motor Vehicle Conflicts Identified Via Active Avoidance Behaviour

Deconstructing Pedestrian Crossing Decision-making in Interactions with Continuous Traffic: an Anthropomorphic Model

A collision-free model on the interaction between pedestrians and cyclists on a shared road

Driving Behavior Analysis for Pedestrian Collision Avoidance Under Emergency Scenarios

Avoidance Behaviors of Pedestrians in a Virtual-Reality-based Experiment

Analysis and Prediction of Pedestrians' Violation Behavior at the Intersection Based on a Markov Chain

Modeling pedestrian flow accounting for collision avoidance during evacuation

Coupling intention and actions of vehicle–pedestrian interaction: A virtual reality experiment study

Understanding and Modeling Collision Avoidance Behavior for Realistic Crowd Simulation

Development of pedestrian collision avoidance strategy based on the fusion of Markov and social force models

Social aspects of collision avoidance: A detailed analysis of two-person groups and individual pedestrians

Identifying factors that contribute to collision avoidance behaviours while walking in a natural environment

Ecological Data Reveal Imbalances in Collision Avoidance Due to Groups' Social Interaction

Kinetic and Kinematic Features of Pedestrian Avoidance Behavior in Motor Vehicle Conflicts

Simulation Study on Pedestrian Strategy Choice Based on Direction Fuzzy Visual Field.

Improved Social Force Model based on Pedestrian Collision Avoidance Behavior in Counterflow