Example-based learning of codes that statistically encode general image classes is of vital importance for computational vision. Recently non negative matrix factorization (NMF) was suggested to provide image code that was both sparse and localized, in contrast to established non local methods like PCA. In this paper, we adopt and generalize this approach to develop a novel learning framework that allows to efficiently compute sparsity-controlled invariant image codes by a well defined sequence of convex conic programs. Applying the corresponding parameter-free algorithm to various image classes results in semantically relevant and transformation-invariant image representations that are remarkably robust against noise and quantization.


    Zugriff

    Zugriff prüfen

    Verfügbarkeit in meiner Bibliothek prüfen

    Bestellung bei Subito €


    Exportieren, teilen und zitieren



    Titel :

    Learning non-negative sparse image codes by convex programming


    Beteiligte:
    Heiler, M. (Autor:in) / Schnorr, C. (Autor:in)


    Erscheinungsdatum :

    2005-01-01


    Format / Umfang :

    315943 byte





    Medientyp :

    Aufsatz (Konferenz)


    Format :

    Elektronische Ressource


    Sprache :

    Englisch



    Learning Non-Negative Sparse Image Codes by Convex Programming

    Heiler, M. / Schnorr, C. / IEEE | British Library Conference Proceedings | 2005



    Image classification by non-negative sparse coding, correlation constrained low-rank and sparse decomposition

    Zhang, C. / Liu, J. / Liang, C. et al. | British Library Online Contents | 2014


    Convex programming on Riemannian manifolds

    Rapcsák, T. | Springer Verlag | 1986


    Sparse Image Coding Using a 3D Non-Negative Tensor Factorization

    Hazan, T. / Polak, S. / Shashua, A. et al. | British Library Conference Proceedings | 2005