As the pretraining technique is growing in popularity, little work has been done on pretrained learning-based motion prediction methods in autonomous driving. In this paper, we propose a framework to formalize the pretraining task for trajectory prediction of traffic participants. Within our framework, inspired by the random masked model in natural language processing (NLP) and computer vision (CV), objects' positions at random timesteps are masked and then filled in by the learned neural network (NN). By changing the mask profile, our framework can easily switch among a range of motion-related tasks. We show that our proposed pretraining framework is able to deal with noisy inputs and improves the motion prediction accuracy and miss rate, especially for objects occluded over time by evaluating it on Argoverse and NuScenes datasets.


    Zugriff

    Zugriff prüfen

    Verfügbarkeit in meiner Bibliothek prüfen

    Bestellung bei Subito €


    Exportieren, teilen und zitieren



    Titel :

    RMP: A Random Mask Pretrain Framework for Motion Prediction


    Beteiligte:
    Yang, Yi (Autor:in) / Zhang, Qingwen (Autor:in) / Gilles, Thomas (Autor:in) / Batool, Nazre (Autor:in) / Folkesson, John (Autor:in)


    Erscheinungsdatum :

    24.09.2023


    Format / Umfang :

    1255684 byte





    Medientyp :

    Aufsatz (Konferenz)


    Format :

    Elektronische Ressource


    Sprache :

    Englisch



    Statistical Motion Mask and Sliding Registration

    Eiben, B / Tran, EH / Menten, MJ et al. | BASE | 2018

    Freier Zugriff

    Enhancing Motion Prediction by a Cooperative Framework

    Araluce, Javier / Justo, Alberto / Arizala, Asier et al. | IEEE | 2024


    Extreme value prediction of the roll motion under random seas

    Jie ZHU / Zailiang LIU / Yan LIN et al. | DOAJ | 2025

    Freier Zugriff

    A Generic Framework for Motion Prediction in Autonomous Driving

    Karle, Phillip Jonathan | TIBKAT | 2024

    Freier Zugriff

    MOTION PREDICTION DEVICE, MOTION PREDICTION METHOD, AND MOTION PREDICTION PROGRAM

    OKAYAMA KEN | Europäisches Patentamt | 2022

    Freier Zugriff