Multiple object tracking (MOT) remains an open and challenging problem for autonomous vehicles. Existing methods mainly ignore prior information from real traffic scenes. Here, the authors propose a novel MOT algorithm that considers traffic safety for vulnerable road users. The proposed method integrates two attention modules with a novel detection refinement strategy. Since skilled drivers pay more attention to pedestrians and cyclists, the authors employ a saliency detection method to extract scene attention region. Then, a detection refinement strategy achieved a good trade-off between parallel single object trackers and detection results. Channel attention can mine the most useful feature channel for traffic road users. In the end, the authors operate their method on the popular MOT 17 benchmark in comparison with other high-level MOT algorithms. The tracking results show that the proposed dual-attention network achieves the state-of-the-art performance.
Multiple object tracking using a dual-attention network for autonomous driving
IET Intelligent Transport Systems ; 14 , 8 ; 842-848
2020-04-07
7 pages
Article (Journal)
Electronic Resource
English
object detection , high-level MOT algorithms , road safety , object tracking , autonomous driving , parallel single object trackers , traffic road users , traffic safety , novel detection refinement strategy , open problem , autonomous vehicles , traffic scenes , saliency detection method , multiple object tracking , vulnerable road users , channel attention , road traffic , scene attention region , traffic engineering computing , feature extraction , attention modules , popular MOT 17 benchmark , dual-attention network , novel MOT algorithm
Metadata by IET is licensed under CC BY 3.0
Multiple object tracking using a dual‐attention network for autonomous driving
Wiley | 2020
|A Tracking-By-Detection Based 3D Multiple Object Tracking for Autonomous Driving
Springer Verlag | 2022
|A novel Siamese Attention Network for visual object tracking of autonomous vehicles
SAGE Publications | 2021
|SELECTIVE ATTENTION FOR DETECTION AND TRACKING OF ROAD-NETWORKS IN AUTONOMOUS DRIVING
British Library Conference Proceedings | 2013
|