Generalized belief propagation GBP is a region-based belief propagation algorithm which can get good convergence in Markov random fields. However, the computation time is too heavy to use in practical engineering applications. This paper proposes a method to accelerate the efficiency of GBP. A caching technique and chessboard passing strategy are used to speed up algorithm. Then, the direction ...

Stochastic gradient descent (SGD) has achieved great success in training deep neural network, where the gradient is computed through backpropagation. However, the back-propagated values of different layers vary dramatically. This inconsistence of gradient magnitude across different layers renders optimization of deep neural network with a single learning rate problematic. We introduce the back-...

Belief propagation is a message-passing algorithm for performing inference in graphical models that has been implemented in a variety of domains including artificial intelligence, information theory and statistical physics. As we know, BP provides exact results only on a relatively restrictive subclass of problems, when the graphical model is a tree (or more generally if each connected componen...

We investigate an algorithm derived based on the belief propagation method of Pearl [11] applied to the (Min-)Bisection problem under the standard planted solution model (or more precisely the Most Likely Partition problem under the same planted solution model). We first point out that the algorithm (without thresholding) is nothing but the standard power method for computing an eigenvector wit...

We define an algebra of ternary relations for cyclic ordering of 2D orientations, which is a refinement of the CYCORD theory. The algebra (1) contains atomic relations, hence 224 general relations, of which the usual CYCORD relation is a particular relation; and (2) is NP-complete, which is not surprising since the CYCORD theory is. We then provide the following: (1) a constraint propagation al...

We describe a two-stage approach to the use of pixel based neural networks for object detection problems in which the locations of relatively small objects in large pictures must be found. The networks use a squared input eld which is large enough to contain all objects of interest. In the rst stage the network is trained on examples which have been cut out from the large pictures. A back error...

This paper examines the implementation of an automated W-CDMA microcellular deployment and coverage reconfiguration algorithm based on the concepts of Situation Awareness (SA). A deterministic microcellular propagation model (Citrus) is used to provide the detailed site-specific propagation data. An initial network deployment is performed over one square kilometre of central Bristol using the C...

The linear canonical transform provides a mathematical model of paraxial propagation though quadratic phase systems. We review the literature on numerical approximation of this transform, including discretization, sampling, and fast algorithms, and identify key results. We then propose a frequency-division fast linear canonical transform algorithm comparable to the Sande-Tukey fast Fourier tran...

This paper describes a parallel implementation of a texture segmentation algorithm. The algorithm uses a Pearl Bayes Network (PBN) to combine evidence for the location of urban regions in airborne infra-red linescan images and for the location of driveable regions in autonomous land vehicle imagery. A multilevel PBN approach is introduced and followed by an example which is used to illustrate t...

This paper presents a novel intermediate view synthesis method based on adaptive belief propagation (BP) algorithm and view interpolation. First, we raise the accuracy in disparity estimation by introducing an adaptive BP algorithm. Then, image pairs are divided into three kinds of regions. Finally, intermediate view is obtained by applying a new interpolation method to each region. Experimenta...