 |
    |
Non-parametric Higher-Order Random Fields for Image Segmentation
Pablo Márquez-Neila1, Pushmeet Kohli2, Carsten Rother3, and Luis Baumela1
1Universidad Politécnica de Madrid, Spain
p.mneila@upm.es
lbaumela@upm.es
2Microsoft Research Cambridge, UK
pkohli@microsoft.com
3Technische Universität Dresden, Germany
carsten.rother@tu-dresden.de
Abstract. Models defined using higher-order potentials are becoming increasingly popular in computer vision. However, the exact representation of a general higher-order potential defined over many variables is computationally unfeasible. This has led prior works to adopt parametric potentials that can be compactly represented. This paper proposes a non-parametric higher-order model for image labeling problems that uses a patch-based representation of its potentials. We use the transformation scheme of [11, 25] to convert the higher-order potentials to a pair-wise form that can be handled using traditional inference algorithms. This representation is able to capture structure, geometrical and topological information of labels from training data and to provide more precise segmentations. Other tasks such as image denoising and reconstruction are also possible. We evaluate our method on denoising and segmentation problems with synthetic and real images.
Keywords: random fields, biomedical image analysis, higher-order models, image denoising, image segmentation, structured prediction
LNCS 8694, p. 269 ff.
lncs@springer.com
© Springer International Publishing Switzerland 2014