Results 1  10
of
12
Learning lowlevel vision
 International Journal of Computer Vision
, 2000
"... We show a learningbased method for lowlevel vision problems. We setup a Markov network of patches of the image and the underlying scene. A factorization approximation allows us to easily learn the parameters of the Markov network from synthetic examples of image/scene pairs, and to e ciently prop ..."
Abstract

Cited by 468 (25 self)
 Add to MetaCart
We show a learningbased method for lowlevel vision problems. We setup a Markov network of patches of the image and the underlying scene. A factorization approximation allows us to easily learn the parameters of the Markov network from synthetic examples of image/scene pairs, and to e ciently propagate image information. Monte Carlo simulations justify this approximation. We apply this to the \superresolution " problem (estimating high frequency details from a lowresolution image), showing good results. For the motion estimation problem, we show resolution of the aperture problem and llingin arising from application of the same probabilistic machinery.
The Dynamics of Nonlinear Relaxation Labeling Processes
, 1997
"... We present some new results which definitively explain the behavior of the classical, heuristic nonlinear relaxation labeling algorithm of Rosenfeld, Hummel, and Zucker in terms of the HummelZucker consistency theory and dynamical systems theory. In particular, it is shown that, when a certain symm ..."
Abstract

Cited by 31 (10 self)
 Add to MetaCart
We present some new results which definitively explain the behavior of the classical, heuristic nonlinear relaxation labeling algorithm of Rosenfeld, Hummel, and Zucker in terms of the HummelZucker consistency theory and dynamical systems theory. In particular, it is shown that, when a certain symmetry condition is met, the algorithm possesses a Liapunov function which turns out to be (the negative of) a wellknown consistency measure. This follows almost immediately from a powerful result of Baum and Eagon developed in the context of Markov chain theory. Moreover, it is seen that most of the essential dynamical properties of the algorithm are retained when the symmetry restriction is relaxed. These properties are also shown to naturally generalize to higherorder relaxation schemes. Some applications and implications of the presented results are finally outlined.
Continuoustime Relaxation Labeling Processes
, 1998
"... We study the properties of two new relaxation labeling schemes described in terms of differential equations, and hence evolving in countinuous time. This contrasts with the customary approach to defining relaxation labeling algorithms which prefers discrete time. Continuoustime dynamical systems ar ..."
Abstract

Cited by 19 (4 self)
 Add to MetaCart
We study the properties of two new relaxation labeling schemes described in terms of differential equations, and hence evolving in countinuous time. This contrasts with the customary approach to defining relaxation labeling algorithms which prefers discrete time. Continuoustime dynamical systems are particularly attractive because they can be implemented directly in hardware circuitry, and the study of their dynamical properties is simpler and more elegant. They are also more plausible as models of biological visual computation. We prove that the proposed models enjoy exactly the same dynamical properties as the classical relaxation labeling schemes, and show how they are intimately related to Hummel and Zucker's now classical theory of constraint satisfaction. In particular, we prove that, when a certain symmetry condition is met, the dynamical systems' behavior is governed by a Liapunov function which turns out to be (the negative of) a wellknown consistency measure. Moreover, we p...
Autoassociative Learning in Relaxation Labeling Networks
, 1997
"... We address the problem of training relaxation labeling processes, a popular class of parallel iterative procedures widely employed in pattern recognition and computer vision. The approach discussed here is based on a theory of consistency developed by Hummel and Zucker, and contrasts with a recently ..."
Abstract

Cited by 5 (3 self)
 Add to MetaCart
We address the problem of training relaxation labeling processes, a popular class of parallel iterative procedures widely employed in pattern recognition and computer vision. The approach discussed here is based on a theory of consistency developed by Hummel and Zucker, and contrasts with a recently introduced learning strategy which can be regarded as heteroassociative, i.e., what is actually learned is the association between patterns rather than the patterns themselves. The proposed learning model is instead autoassociative and involves making a set of training patterns consistent, in the sense rigorously defined by Hummel and Zucker; this implies that they become local attractors of the relaxation labeling dynamical system. The learning problem is formulated in terms of solving a system of linear inequalities, and a straightforward iterative algorithm is presented to accomplish this. The learning model described here allows one to view the relaxation labeling process as a kind of ...
Image Segmentation and Labeling Using the Polya Urn Model
 Robustness of HMM and ANN Speech Recognition Algorithms. Proceedings of the International Conference on Spoken Language Processing
, 1990
"... We propose a segmentation method based on Polya’s urn model for contagious phenomena. A preliminary segmentation yields the initial composition of an urn representing the pixel. The resulting urns are then subjected to a modified urn sampling scheme mimicking the development of an infection to yiel ..."
Abstract

Cited by 3 (0 self)
 Add to MetaCart
We propose a segmentation method based on Polya’s urn model for contagious phenomena. A preliminary segmentation yields the initial composition of an urn representing the pixel. The resulting urns are then subjected to a modified urn sampling scheme mimicking the development of an infection to yield a segmentation of the image into homogeneous regions. This process is implemented using contagion urn processes and generalizes Polya’s scheme by allowing spatial interactions. The composition of the urns is iteratively updated by assuming a spatial Markovian relationship between neighboring pixel labels. The asymptotic behavior of this process is examined and comparisons with simulated annealing and relaxation labeling are presented. Examples of the application of this scheme to the segmentation of synthetic texture images, ultrawideband synthetic aperture radar (UWB SAR) images and magnetic resonance images (MRI) are provided.
Separating Overlapped Fingerprints
"... Abstract—Fingerprint images generally contain either a single fingerprint (e.g., rolled images) or a set of nonoverlapped fingerprints (e.g., slap fingerprints). However, there are situations where several fingerprints overlap on top of each other. Such situations are frequently encountered when lat ..."
Abstract

Cited by 2 (2 self)
 Add to MetaCart
Abstract—Fingerprint images generally contain either a single fingerprint (e.g., rolled images) or a set of nonoverlapped fingerprints (e.g., slap fingerprints). However, there are situations where several fingerprints overlap on top of each other. Such situations are frequently encountered when latent (partial) fingerprints are lifted from crime scenes or residue fingerprints are left on fingerprint sensors. Overlapped fingerprints constitute a serious challenge to existing fingerprint recognition algorithms, since these algorithms are designed under the assumption that fingerprints have been properly segmented. In this paper, a novel algorithm is proposed to separate overlapped fingerprints into component or individual fingerprints. The basic idea is to first estimate the orientation field of the given image with overlapped fingerprints and then separate it into component orientation fields using a relaxation labeling technique. We also propose an algorithm to utilize fingerprint singularity information to further improve the separation performance. Experimental results indicate that the algorithm leads to good separation of overlapped fingerprints that leads to a significant improvement in the matching accuracy. Index Terms—Fingerprint matching, fingerprint separation, latent fingerprints, orientation field, overlapped fingerprints, relaxation labeling, singularity. I.
Supervised Segmentation by Iterated Contextual Pixel Classification
 In Proceedings 16th International Conference on Pattern Recognition
, 2002
"... We propose a general iterative contextual pixel classifier for supervised image segmentation. The iterative procedure is statistically wellfounded, and can be considered a variation on the iterated conditional modes (ICM) of Besag. Having an initial segmentation, the algorithm iteratively updates i ..."
Abstract

Cited by 2 (2 self)
 Add to MetaCart
We propose a general iterative contextual pixel classifier for supervised image segmentation. The iterative procedure is statistically wellfounded, and can be considered a variation on the iterated conditional modes (ICM) of Besag. Having an initial segmentation, the algorithm iteratively updates it by reclassifying every pixel, based on the original features and, additionally, contextual information. This contextual information consists of the class labels of pixels in the neighborhood of the pixel to be reclassified. Three essential differences with the original ICM are: 1) Our update step is merely based on a classification result, hence avoiding the explicit calculation of conditional probabilities; 2) The clique formalism of the Markov random field framework is not required; 3) No assumption is made w.r.t. the conditional independence of the observed pixel values given the segmented image. The important consequence of Properties 1 and 2 is that one can easily incorporate common pattern recognition tools in our segmentation algorithm. Examples are different classifierse.g. Fisher linear discriminant, nearestneighbor classifier, or support vector machinesand dimension reduction techniques like LDA, or PCA. We experimentally compare a specific instance of our general method to pixel classification, using simulated data and chest radiographs, and show that the former outperforms the latter.
General Geometric Good Continuation: From Taylor to Laplace via Level Sets
"... that parts often group in particular ways to form coherent wholes. Perceptual integration of edges, for example, involves orientation good continuation, a property which has been exploited computationally very extensively. But more general localglobal relationships, such as for shading or color, ha ..."
Abstract
 Add to MetaCart
that parts often group in particular ways to form coherent wholes. Perceptual integration of edges, for example, involves orientation good continuation, a property which has been exploited computationally very extensively. But more general localglobal relationships, such as for shading or color, have been elusive. While Taylor’s Theorem suggests certain modeling and smoothness criteria, the consideration of level set geometry indicates a different approach. Using such first principles we derive, for the first time, a generalization of good continuation to all those visual structures that can be abstracted as scalar functions over the image plane. Based on second order differential constraints that reflect good continuation, our analysis leads to a unique class of harmonic models and a cooperative algorithm for structure inference. Among the different applications of good continuation, here we apply these results to the denoising of shading and intensity distributions and demonstrate how our approach eliminates spurious measurements while preserving both singularities and regular structure, a property that facilitates higher level processes which depend so critically on both of these classes of visual structures.
A Comparison of Stochastic Optimization Techniques for Image Segmentation
"... Image segmentation denotes a process by which a raw input image is partitioned into nonoverlapping regions such that each region is homogeneous and the union of any two adjacent regions is heterogenous. A segmented image is considered to be the highest domainindependent abstraction of an input imag ..."
Abstract
 Add to MetaCart
Image segmentation denotes a process by which a raw input image is partitioned into nonoverlapping regions such that each region is homogeneous and the union of any two adjacent regions is heterogenous. A segmented image is considered to be the highest domainindependent abstraction of an input image. In this paper, the image segmentation problem is treated as one of combinatorial optimization. A cost function which incorporates both, edge information and region grayscale variances is defined. The cost function is shown to be multivariate with several local minima. Three stochastic optimization techniques, namely, simulated annealing Ž SA., microcanonical annealing Ž MCA., and the random cost algorithm Ž RCA. are investigated and compared in the context of minimization of the aforementioned cost function for image segmentation. Experimental results on grayscale images are presented. � 2000 John Wiley & Sons, Inc. I.
Probabilistic Relaxation as an Optimizer 613
"... Probabilistic Relaxation with product support has been shown to have advantages over 'traditional ' probabilistic relaxation. However it is less well understood in the sense that a cost function is not known. In this paper we present a cost function. This greatly improves our understanding of probab ..."
Abstract
 Add to MetaCart
Probabilistic Relaxation with product support has been shown to have advantages over 'traditional ' probabilistic relaxation. However it is less well understood in the sense that a cost function is not known. In this paper we present a cost function. This greatly improves our understanding of probabilistic relaxation with product support, and also leads us to propose a new class of probabilistic relaxation algorithms. We investigate two applications. 1