Floors are Flat - Leveraging Semantics for Real-Time Surface Normal Prediction

    ICCV Workshops, pp. 4065-4074, 2019.

    Cited by: 0|Bibtex|Views4|Links
    EI

    Abstract:

    We propose 4 insights that help to significantly improve the performance of deep learning models that predict surface normals and semantic labels from a single RGB image. These insights are: (1) denoise the "ground truth" surface normals in the training set to ensure consistency with the semantic labels; (2) concurrently train on a mix ...More

    Code:

    Data:

    Your rating :
    0

     

    Tags
    Comments