Minimax entropy and learning by diffusion

A system of coupled differential equations is formulated which learns priors for modelling "preattentive" textures. It is derived from an energy functional consisting of a linear combination of a large number of terms corresponding to the features that the system is capable of learning. The system learns the parameters associated with each feature by applying gradient ascent to the log-likelihood function. Update of each parameter is thus governed by the residual with respect to the corresponding feature. A feature residual is computed from its observed value and value generated by the system. The latter is calculated from a synthesized sample image, which is generated by means of a reaction-diffusion equation obtained by applying gradient descent to the energy functional.

[1]  MumfordDavid,et al.  Filters, Random Fields and Maximum Entropy (FRAME) , 1998 .

[2]  Andrea Braides,et al.  Non-local approximation of the Mumford-Shah functional , 1997 .

[3]  Jayant Shah,et al.  Parameter estimation, multiscale representation and algorithms for energy-minimizing segmentations , 1990, [1990] Proceedings. 10th International Conference on Pattern Recognition.

[4]  Andrew Blake,et al.  Visual Reconstruction , 1987, Deep Learning for EEG-Based Brain–Computer Interfaces.

[5]  Donald Geman,et al.  Stochastic relaxation, Gibbs distributions, and the Bayesian restoration of images , 1984 .

[6]  A.S. Sherstinsky,et al.  M-lattice: from morphogenesis to image processing , 1996, IEEE Trans. Image Process..

[7]  Jitendra Malik,et al.  Normalized cuts and image segmentation , 1997, Proceedings of IEEE Computer Society Conference on Computer Vision and Pattern Recognition.

[8]  Jayant Shah,et al.  A common framework for curve evolution, segmentation and anisotropic diffusion , 1996, Proceedings CVPR IEEE Computer Society Conference on Computer Vision and Pattern Recognition.

[9]  Song-Chun Zhu,et al.  GRADE: Gibbs reaction and diffusion equations , 1998, Sixth International Conference on Computer Vision (IEEE Cat. No.98CH36271).

[10]  Jitendra Malik,et al.  Scale-Space and Edge Detection Using Anisotropic Diffusion , 1990, IEEE Trans. Pattern Anal. Mach. Intell..

[11]  R. Jensen.  Uniqueness of Lipschitz extensions: Minimizing the sup norm of the gradient , 1993 .

[12]  Donald Geman,et al.  Stochastic Relaxation, Gibbs Distributions, and the Bayesian Restoration of Images , 1984, IEEE Transactions on Pattern Analysis and Machine Intelligence.