Autonomous driving and its widespread adoption have long held tremendous promise. Nevertheless, without a trustworthy and thorough testing procedure, not only does the industry struggle to mass-produce autonomous vehicles (AV), but neither the general public nor policymakers are convinced to accept the innovations. Generating safety-critical scenarios that present significant challenges to AV is an essential first step in testing. Real-world datasets include naturalistic but overly safe driving behaviors, whereas simulation would allow for unrestricted exploration of diverse and aggressive traffic scenarios. Conversely, higher-dimensional searching space in simulation disables efficient scenario generation without real-world data distribution as implicit constraints. In order to marry the benefits of both, it seems appealing to learn to generate scenarios from both offline real-world and online simulation data simultaneously. Therefore, we tailor a Reversely Regularized Hybrid Offline-and-Online ((Re)2H2O) Reinforcement Learning recipe to additionally penalize Q-values on real-world data and reward Q-values on simulated data, which ensures the generated scenarios are both varied and adversarial. Through extensive experiments, our solution proves to produce more risky scenarios than competitive baselines and it can generalize to work with various autonomous driving models. In addition, these generated scenarios are also corroborated to be capable of fine-tuning AV performance.


    Zugriff

    Zugriff prüfen

    Verfügbarkeit in meiner Bibliothek prüfen

    Bestellung bei Subito €


    Exportieren, teilen und zitieren



    Titel :

    (Re)2H2O: Autonomous Driving Scenario Generation via Reversely Regularized Hybrid Offline-and-Online Reinforcement Learning


    Beteiligte:
    Niu, Haoyi (Autor:in) / Ren, Kun (Autor:in) / Xu, Yizhou (Autor:in) / Yang, Ziyuan (Autor:in) / Lin, Yichen (Autor:in) / Zhang, Yi (Autor:in) / Hu, Jianming (Autor:in)


    Erscheinungsdatum :

    04.06.2023


    Format / Umfang :

    1054153 byte





    Medientyp :

    Aufsatz (Konferenz)


    Format :

    Elektronische Ressource


    Sprache :

    Englisch



    Offline Reinforcement Learning for Autonomous Driving with Real World Driving Data

    Fang, Xing / Zhang, Qichao / Gao, Yinfeng et al. | IEEE | 2022


    OFFLINE AND ONLINE COMPENSATION FOR CONTROL IN AUTONOMOUS DRIVING

    QU SHEN / LIU JINGXUAN / KOLARIC PATRIK et al. | Europäisches Patentamt | 2025

    Freier Zugriff

    AUTONOMOUS VEHICLE TESTING OPTIMIZATION USING OFFLINE REINFORCEMENT LEARNING

    LEE RITCHIE | Europäisches Patentamt | 2024

    Freier Zugriff


    Automatic driving decision-making method based on offline and online fusion reinforcement learning

    ZHANG KAIXIN / YU SIJIU | Europäisches Patentamt | 2025

    Freier Zugriff