Reconstructing dynamic traffic scenes has a wide range of applications in the development of modern autonomous driving systems. Recently, novel-view synthesis techniques, such as Neural Radiance Field (NeRF) and 3D Gaussian Splatting (3D GS), have emerged as promising paradigms for the re-construction of 3D scenes. However, previous works in this area highly rely on LiDAR to provide accurate geometric prior and motion cues across frames to reconstruct dynamic objects, hindering the use of overwhelming vision data from mass-production vehicles. In this paper, we propose a novel Vision-Centric reconstruction framework based on 3D GS, VC-Gaussian, which allows high-quality novel-view synthesis and dynamic scene reconstruction for autonomous driving. A composite Gaussian model is designed to represent background and foreground objects separately. To facilitate the initialization and optimization process of Gaussians without LiDAR, we leverage easy-to-obtain monocular geometric prior including metric depth and normal. Experimental results on the real autonomous driving dataset demonstrate that our method out-performs other reconstruction methods with monocular vision inputs and even is competitive with LiDAR-based methods.


    Zugriff

    Zugriff prüfen

    Verfügbarkeit in meiner Bibliothek prüfen

    Bestellung bei Subito €


    Exportieren, teilen und zitieren



    Titel :

    VC-Gaussian: Vision-Centric Gaussian Splatting for Dynamic Autonomous Driving Scenes


    Beteiligte:
    Li, Jiusi (Autor:in) / Wen, Tuopu (Autor:in) / Jiang, Kun (Autor:in) / Miao, Jinyu (Autor:in) / Shi, Yining (Autor:in) / Zhao, Xuhe (Autor:in) / Fan, Zhigang (Autor:in) / Yang, Diange (Autor:in)


    Erscheinungsdatum :

    24.09.2024


    Format / Umfang :

    1460324 byte





    Medientyp :

    Aufsatz (Konferenz)


    Format :

    Elektronische Ressource


    Sprache :

    Englisch



    Characterizing Satellite Geometry via Accelerated 3D Gaussian Splatting

    Nguyen, Van Minh / Sandidge, Emma / Mahendrakar, Trupti et al. | ArXiv | 2024

    Freier Zugriff

    DMP: Difference-Guided Motion Prediction for Vision-Centric Autonomous Driving

    Chen, Yanyan / Lin, Chunmian / Duan, Xuting et al. | IEEE | 2025


    Spatiotemporal Gaussian mixture model to detect moving objects in dynamic scenes

    Zhang, W. / Fang, X. / Yang, X. et al. | British Library Online Contents | 2007


    Product Correlator Performance for Gaussian Random Scenes

    Webber, Robert F. / Delashmit, Walter H. | IEEE | 1974


    Reachability Estimation in Dynamic Driving Scenes for Autonomous Vehicles

    Medina-Lee, Juan Felipe / Artuñedo, Antonio / Godoy, Jorge et al. | BASE | 2020

    Freier Zugriff