The rise of self-driving cars (SDCs) has sparked concerns about the safety of vulnerable road users, such as pedestrians. While autonomous vehicles (AVs) have the potential to reduce accidents caused by human error, one major challenge lies in accurately detecting and understanding pedestrians' intentions' especially in unprecedented events such as, detecting a pedestrian on a high-way. Additionally, pedestrians often struggle to anticipate AV behavior or effectively communicate with them. To address these issues, this study aims to identify the expedient communication modality to be used by AVs to communicate their intent to pedestrians. To this end, we developed a virtual reality (VR) simulation of a virtual high-way. In our VR simulation, we developed (1) a virtual AV that is trained using object detection model that utilizes posture analysis to accurately detect pedes-trians' intentions, and (2) two different communication methods (visual and verbal) between AVs and pedestrians. To achieve our aim, we conducted a user study (N = 24) to investigate the effect of four different communication modalities on user's trust of AVs. Our results showed that pedestrians expressed more trust toward AVs that communicate their intentions irrespective of the adopted modality, and preferred that AVs use a combination of visual and verbal modalities. We believe that our findings contribute to developing more effective Pedestrian-AV communication systems. We envision that our work will be a step toward integrating AVs in future smart cities.


    Zugriff

    Zugriff prüfen

    Verfügbarkeit in meiner Bibliothek prüfen

    Bestellung bei Subito €


    Exportieren, teilen und zitieren



    Titel :

    Building Trust in AV-to-Pedestrian Interaction: Exploring the Influence of Communication Modality through a Virtual Reality Study


    Beteiligte:
    Yassien, Amal (Autor:in) / Hammouda, Somaya (Autor:in) / Sharaf, Nada (Autor:in) / Abdennadher, Slim (Autor:in)


    Erscheinungsdatum :

    22.10.2023


    Format / Umfang :

    2392430 byte




    Medientyp :

    Aufsatz (Konferenz)


    Format :

    Elektronische Ressource


    Sprache :

    Englisch



    Pedestrian recognition through different cross-modality deep learning methods

    Pop, Danut Ovidiu / Rogozan, Alexandrina / Nashashibi, Fawzi et al. | IEEE | 2017



    A Review of Virtual Reality Studies on Autonomous Vehicle--Pedestrian Interaction

    Tran, Tram Thi Minh / Parker, Callum / Tomitsch, Martin | ArXiv | 2024

    Freier Zugriff

    Incremental Cross-Modality deep learning for pedestrian recognition

    Pop, Danut Ovidiu / Rogozan, Alexandrina / Nashashibi, Fawzi et al. | IEEE | 2017


    Incremental Cross-Modality Deep Learning for Pedestrian Recognition

    Pop, Danut Ovidiu / Rogozan, Alexandria / Nashashibi, Fawzi et al. | British Library Conference Proceedings | 2017