Modern models that perform system-critical tasks such as segmentation and localization exhibit good performance and robustness under ideal conditions (i.e. daytime, overcast) but performance degrades quickly and often catastrophically when input conditions change. In this work, we present a domain adaptation system that uses light-weight input adapters to pre-processes input images, irrespective of their appearance, in a way that makes them compatible with off-the-shelf computer vision tasks that are trained only on inputs with ideal conditions. No fine-tuning is performed on the off-the-shelf models, and the system is capable of incrementally training new input adapters in a self-supervised fashion, using the computer vision tasks as supervisors, when the input domain differs significantly from previously seen domains. We report large improvements in semantic segmentation and topological localization performance on two popular datasets, RobotCar and BDD.


    Access

    Check access

    Check availability in my library

    Order at Subito €


    Export, share and cite



    Title :

    Don’t Worry About the Weather: Unsupervised Condition-Dependent Domain Adaptation


    Contributors:
    Porav, Horia (author) / Bruls, Tom (author) / Newman, Paul (author)


    Publication date :

    2019-10-01


    Size :

    2763410 byte




    Type of media :

    Conference paper


    Type of material :

    Electronic Resource


    Language :

    English




    Should Optimal Designers Worry About Consideration?

    Ross Morrow, W | Online Contents | 2015


    Should Designers Worry About Market Systems?

    Shiau, Ching-Shin Norman | Online Contents | 2009


    Unsupervised Evaluation of Lidar Domain Adaptation

    Hubschneider, Christian / Roesler, Simon / Zollner, J. Marius | IEEE | 2020


    Why Should We Worry About the Railways?

    Bradshaw, B. | British Library Online Contents | 1996