Abstract--A crucial issue in designing learning machines is to select the correct model parameters. When the number of available samples is small, theoretical sample-based generali...
Background: Traditional algorithms for hidden Markov model decoding seek to maximize either the probability of a state path or the number of positions of a sequence assigned to th...
We propose a timed process calculus for wireless systems exposed to communication collisions. The operational semantics of our calculus is given in terms of a labelled transition ...
Abstract. Bottom-up approaches, which rely mainly on continuity principles, are often insufficient to form accurate segments in natural images. In order to improve performance, rec...
Image segmentation plays an important role in computer vision and image analysis. In this paper, the segmentation problem is formulated as a labeling problem under a probability m...