3D detection of traffic management objects, such as traffic lights and road signs, is vital for self-driving cars, particularly for address-to-address navigation where vehicles encounter numerous intersections with these static objects. This paper introduces a novel method for automatically generating accurate and temporally consistent 3D bounding box annotations for traffic lights and signs, effective up to a range of 200 meters. These annotations are suitable for training real-time models used in self-driving cars, which need a large amount of training data. The proposed method relies only on RGB images with 2D bounding boxes of traffic management objects, which can be automatically obtained using an off-the-shelf image-space detector neural network, along with GNSS/INS data, eliminating the need for LiDAR point cloud data.


    Access

    Download


    Export, share and cite



    Title :

    Accurate Automatic 3D Annotation of Traffic Lights and Signs for Autonomous Driving


    Contributors:

    Publication date :

    2024


    Remarks:

    Accepted at the 2nd Workshop on Vision-Centric Autonomous Driving (VCAD) as part of ECCV 2024


    Type of media :

    Preprint


    Type of material :

    Electronic Resource


    Language :

    English




    Supporting device for traffic lights and signs

    KIM MAN SEOK / PARK GEN TAE / YONG CHAEMYEONG | European Patent Office | 2021

    Free access

    Traffic lights system based on RFID for autonomous driving vehicle

    Ali, AbduladhemAbdulkareem / Hussein, Hussein Alaa | IEEE | 2017


    Point Cloud Automatic Annotation Framework for Autonomous Driving

    Zhao, Chaoran / Peng, Bo / Azumi, Takuya | IEEE | 2024


    Automatic traffic signal lights

    Hurley, J.F. | Engineering Index Backfile | 1928