Self-driving cars increasingly rely on deep neural networks to achieve human-like driving. However, the opacity of such black-box motion planners makes it challenging for the human behind the wheel to accurately anticipate when they will fail, with potentially catastrophic consequences. Here, we introduce concept-wrapper network (i.e., CW-Net), a method for explaining the behavior of black-box motion planners by grounding their reasoning in human-interpretable concepts. We deploy CW-Net on a real self-driving car and show that the resulting explanations refine the human driver's mental model of the car, allowing them to better predict its behavior and adjust their own behavior accordingly. Unlike previous work using toy domains or simulations, our study presents the first real-world demonstration of how to build authentic autonomous vehicles (AVs) that give interpretable, causally faithful explanations for their decisions, without sacrificing performance. We anticipate our method could be applied to other safety-critical systems with a human in the loop, such as autonomous drones and robotic surgeons. Overall, our study suggests a pathway to explainability for autonomous agents as a whole, which can help make them more transparent, their deployment safer, and their usage more ethical.


    Access

    Download


    Export, share and cite



    Title :

    Explainable deep learning improves human mental models of self-driving cars


    Contributors:

    Publication date :

    2024


    Remarks:

    * - equal contribution


    Type of media :

    Preprint


    Type of material :

    Electronic Resource


    Language :

    English




    Deep Learning Based Self Driving Cars Using Computer Vision

    Bhaggiaraj, S. / Priyadharsini, M. / Karuppasamy, K. et al. | IEEE | 2023


    Deep Learning for Self-Driving Cars: Chances and Challenges

    Rao, Qing / Frtunikj, Jelena | IEEE | 2018


    Predicting Steering Actions for Self-Driving Cars Through Deep Learning

    Ou, Chaojie / Bedawi, Safaa Mahmoud / Koesdwiady, Arief B. et al. | IEEE | 2018



    SELF DRIVING CARS

    KIM SI HYEONG / HAN YURIM / YEON HEE LEE et al. | European Patent Office | 2022

    Free access