Most current local feature detectors/descriptors implicitly assume that the scene is (locally) planar, an assumption that is violated at surface discontinuities. We show that this restriction is, at least in theory, unnecessary, as one can construct local features that are viewpoint-invariant for generic non-planar scenes. However, we show that any such feature necessarily sacrifices shape information, in the sense of being non shape-discriminative. Finally, we show that if viewpoint is factored out as part of the matching process, rather than explicitly in the representation, then shape is discriminative indeed. We illustrate our theoretical results empirically by showing that, even for simple scenes, current affine descriptors fail where even a naive 3-D viewpoint invariant succeeds in matching.


    Access

    Check access

    Check availability in my library

    Order at Subito €


    Export, share and cite



    Title :

    Features for recognition: viewpoint invariance for non-planar scenes


    Contributors:
    Vedaldi, A. (author) / Soatto, S. (author)


    Publication date :

    2005-01-01


    Size :

    520631 byte





    Type of media :

    Conference paper


    Type of material :

    Electronic Resource


    Language :

    English



    Features for Recognition: Viewpoint Invariance for Non-Planar Scenes

    Vedaldi, A. / Soatto, S. / IEEE | British Library Conference Proceedings | 2005




    Object recognition using multiple view invariance based on complex features

    Kuno, Y. / Takae, O. / Takahashi, T. et al. | IEEE | 1996


    Object Recognition Using Multiple View Invariance Based on Complex Features

    Kuno, Y. / Takae, O. / Takahashi, T. et al. | British Library Conference Proceedings | 1996