Risk-aware Trajectory Prediction by Incorporating Spatio-temporal Traffic Interaction Analysis

TL;DR

This paper introduces a risk-aware trajectory prediction method that incorporates spatio-temporal traffic interaction analysis to improve prediction accuracy in high-risk scenarios for autonomous robots.

Contribution

It presents a novel approach using location-based and speed-based re-weighting techniques to enhance trajectory predictions in dangerous situations.

Findings

Improved most-likely FDE and KDE metrics.

Enhanced prediction accuracy for high-speed vehicles.

Better performance in high-risk locations.

Abstract

To operate in open-ended environments where humans interact in complex, diverse ways, autonomous robots must learn to predict their behaviour, especially when that behavior is potentially dangerous to other agents or to the robot. However, reducing the risk of accidents requires prior knowledge of where potential collisions may occur and how. Therefore, we propose to gain this information by analyzing locations and speeds that commonly correspond to high-risk interactions within the dataset, and use it within training to generate better predictions in high risk situations. Through these location-based and speed-based re-weighting techniques, we achieve improved overall performance, as measured by most-likely FDE and KDE, as well as improved performance on high-speed vehicles, and vehicles within high-risk locations. 2023 IEEE International Conference on Robotics and Automation (ICRA)