The use of genotypic populations is necessary for adaptation in Evolutionary Algorithms. We use a technique called form-invariant commutation to study the immediate effect of evol...
In this paper we present a novel class-based segmentation method, which is guided by a stored representation of the shape of objects within a general class (such as horse images). ...
The goal of a threshold query is to detect all objects whose score exceeds a given threshold. This type of query is used in many settings, such as data mining, event triggering, a...
Guy Sagy, Daniel Keren, Izchak Sharfman, Assaf Sch...
Many problems in machine learning and statistics can be formulated as (generalized) eigenproblems. In terms of the associated optimization problem, computing linear eigenvectors a...
We describe an energy minimization algorithm for functions defined on 4-connected lattices, of the type usually encountered in problems involving images. Such functions are often ...