Trajectory Prediction for Autonomous Driving Using a Transformer Network

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Li, Zhenning, Yu, Hao
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866929256723906560
author Li, Zhenning
Yu, Hao
author_facet Li, Zhenning
Yu, Hao
contents Predicting the trajectories of surrounding agents is still considered one of the most challenging tasks for autonomous driving. In this paper, we introduce a multi-modal trajectory prediction framework based on the transformer network. The semantic maps of each agent are used as inputs to convolutional networks to automatically derive relevant contextual information. A novel auxiliary loss that penalizes unfeasible off-road predictions is also proposed in this study. Experiments on the Lyft l5kit dataset show that the proposed model achieves state-of-the-art performance, substantially improving the accuracy and feasibility of the prediction outcomes.
format Preprint
id arxiv_https___arxiv_org_abs_2402_16501
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Trajectory Prediction for Autonomous Driving Using a Transformer Network
Li, Zhenning
Yu, Hao
Robotics
Predicting the trajectories of surrounding agents is still considered one of the most challenging tasks for autonomous driving. In this paper, we introduce a multi-modal trajectory prediction framework based on the transformer network. The semantic maps of each agent are used as inputs to convolutional networks to automatically derive relevant contextual information. A novel auxiliary loss that penalizes unfeasible off-road predictions is also proposed in this study. Experiments on the Lyft l5kit dataset show that the proposed model achieves state-of-the-art performance, substantially improving the accuracy and feasibility of the prediction outcomes.
title Trajectory Prediction for Autonomous Driving Using a Transformer Network
topic Robotics
url https://arxiv.org/abs/2402.16501