Abstract:In the future driverless scenario, pedestrian-vehicle conflict is an unavoidable traffic problem at unsignalized mid-block crosswalks. Autonomous vehicles are obviously impossible to make eye contact with pedestrians the way a human driver can. Therefore, there is an urgent need for autonomous vehicles to achieve accurate predictions of pedestrian crossing behavior. In order to better solve the problem of high uncertainty of pedestrian crossing behavior, in this paper, a modeling framework combines Maximum Entropy Deep Inverse Reinforcement Learning (Deep MEIRL) and reinforcement learning is employed to predict pedestrian crossing behaviors. The dataset of drone-based video footage is collected in Xi’an (China) to train and validate the model. The trajectory dataset, extracted by computer vision algorithm, is implemented to derive state features of pedestrian behavior, including pedestrian to target area distance, vehicle type, vehicle speed, lateral/longitudinal distances, pedestrian lateral/longitudinal velocities. The results reveal that Deep MEIRL performs better than a baseline model MEIRL at micro-scales such as predicting pedestrian trajectories and evasive actions. Specifically, in predicting pedestrian trajectories, the Deep MEIRL outperforms the MEIRL by 33.09% and 15.16% on the basis of the MAE and HD, respectively. Meanwhile, the Deep MEIRL is 28.7% and 17.6% respectively more accurate than the MEIRL in predicting pedestrian evasive actions on lateral and longitudinal directions. Furthermore, we also found that there is heterogeneity in pedestrian crossing behavior when interacting with different vehicles. This research can contribute to a critical step toward addressing the safe and efficient movement at unsignalized mid-block crosswalks for autonomous vehicles.

A Passing Predicting Model for Autonomous Vehicles at a Mixed-Driving Signaled Intersection

Short-term Traffic Flow Prediction at Isolated Intersections Based on Parallel Multi-Task Learning

Analysis and Modeled Design of One State-Driven Autonomous Passing-Through Algorithm for Driverless Vehicles at Intersections

Vehicle Motion Prediction at Intersections Based on the Turning Intention and Prior Trajectories Model

Driving Intention Recognition and Speed Prediction at Complex Urban Intersections Considering Traffic Environment

A Decision-Making Model for Autonomous Vehicles at Urban Intersections Based on Conflict Resolution

A Study on Autonomous Intersection Management: Planning-Based Strategy Improved by Convolutional Neural Network

Trajectories Prediction of Surrounding Vehicles at Urban Intersections

Modeling of Complex Traffic Interaction Under Uncontrolled Intersection Scenario

Autonomous and Connected Vehicles: The Capacity of Mixed Traffic Flow at Signalized Intersection with the ACDA-MTD Model

A hybrid deep learning model for urban expressway lane-level mixed traffic flow prediction

Left-turn Vehicles Crossing Behavior Prediction and Modeling at Mixed Traffic Flow Intersection.

A Scenario-Adaptive Driving Behavior Prediction Approach to Urban Autonomous Driving

A Hybrid Approach for Turning Intention Prediction Based on Time Series Forecasting and Deep Learning

Predicting Pedestrian Crossing Behavior at Unsignalized Mid-Block Crosswalks Using Maximum Entropy Deep Inverse Reinforcement Learning

Interaction-Aware Personalized Trajectory Prediction for Traffic Participant Based on Interactive Multiple Model

Multi-Intersection Traffic Optimisation: A Benchmark Dataset and a Strong Baseline

Transferable Pedestrian Motion Prediction Models at Intersections

Modeling Vehicle-Pedestrian Encountering Risks in the Natural Driving Environment Using Machine Learning Algorithms

A Decision-Making Model of Left-Turn Behavior for Autonomous Vehicles at Urban Intersections Using a Single-Vehicle Scenario

A novel model for real-time risk evaluation of vehicle-pedestrian interactions at intersections