In autonomous driving, accurate 3D lane detection using monocular cameras is important for downstream tasks. Recent CNN and Transformer approaches usually apply a two-stage model design. The first stage transforms the image feature from a front image into a bird’s-eye-view (BEV) representation. Subsequently, a sub-network processes the BEV feature to generate the 3D detection results. However, these approaches heavily rely on a challenging image feature transformation module from a perspective view to a BEV representation. In our work, we present CurveFormer++, a single-stage Transformer-based method that does not require the view transform module and directly infers 3D lane results from the perspective image features. Specifically, our approach models the 3D lane detection task as a curve propagation problem, where each lane is represented by a curve query with a dynamic and ordered anchor point set. By employing a Transformer decoder, the model can iteratively refine the 3D lane results. A curve cross-attention module is introduced to calculate similarities between image features and curve queries. To handle varying lane lengths, we employ context sampling and anchor point restriction techniques to compute more relevant image features. Furthermore, we apply a temporal fusion module that incorporates selected informative sparse curve queries and their corresponding anchor point sets to leverage historical information. In the experiments, we evaluate our approach on two publicly real-world datasets. The results demonstrate that our method provides outstanding performance compared with both CNN and Transformer based methods. We also conduct ablation studies to analyze the impact of each component.


    Zugriff

    Zugriff prüfen

    Verfügbarkeit in meiner Bibliothek prüfen

    Bestellung bei Subito €


    Exportieren, teilen und zitieren



    Titel :

    CurveFormer++: 3D Lane Detection by Curve Propagation With Temporal Curve Queries and Attention


    Beteiligte:
    Bai, Yifeng (Autor:in) / Chen, Zhirong (Autor:in) / Liang, Pengpeng (Autor:in) / Song, Bo (Autor:in) / Cheng, Erkang (Autor:in)


    Erscheinungsdatum :

    01.06.2025


    Format / Umfang :

    2768111 byte




    Medientyp :

    Aufsatz (Zeitschrift)


    Format :

    Elektronische Ressource


    Sprache :

    Englisch



    Robust curve detection by temporal geodesics

    Gutfinger, D. / Nishimura, R. / Doi, H. et al. | IEEE | 1990


    Research on lane detection algorithm for large curvature curve

    Tian, Shihe / Zhang, Zhian / Huang, Xuegong | SPIE | 2023


    DBNet: A Curve-Based Dynamic Association Framework for Lane Detection

    Dai, Xinguang / Xie, Jun / Zhang, Guoxin et al. | IEEE | 2024


    Lane line preparation method and device for curve

    CHEN DI | Europäisches Patentamt | 2023

    Freier Zugriff

    Adaptive Lane Departure and Curve Negotiation Warning System

    Sentouh, C. / Glaser, S. / Mannar, S. et al. | British Library Conference Proceedings | 2006