MapsTP: HD Map Images Based Multimodal Trajectory Prediction for Automated Vehicles

TL;DR

This paper introduces a multimodal trajectory prediction method for autonomous vehicles that leverages high-definition map images and sensor data to generate and select the most probable future trajectories, enhancing prediction robustness.

Contribution

The novel approach combines HD map images with IMU sensor data using a ResNet-50 feature extractor and a temporal probabilistic network for improved trajectory prediction accuracy.

Findings

Enhanced trajectory prediction accuracy in urban scenarios

Robustness against unpredictable vehicle and pedestrian behaviors

Effective integration of HD map images and sensor data

Abstract

Predicting ego vehicle trajectories remains a critical challenge, especially in urban and dense areas due to the unpredictable behaviours of other vehicles and pedestrians. Multimodal trajectory prediction enhances decision-making by considering multiple possible future trajectories based on diverse sources of environmental data. In this approach, we leverage ResNet-50 to extract image features from high-definition map data and use IMU sensor data to calculate speed, acceleration, and yaw rate. A temporal probabilistic network is employed to compute potential trajectories, selecting the most accurate and highly probable trajectory paths. This method integrates HD map data to improve the robustness and reliability of trajectory predictions for autonomous vehicles.