We present a new method for blind document bleed through removal based on separate Markov Random Field (MRF) regularization for the recto and for the verso side, where separate pri...
This paper introduces a new edge length heuristic that finds a graph layout where the edge lengths are proportional to the weights on the graph edges. The heuristic can be used in...
ABSTRACT. Upon the discovery of power laws [8, 16, 30], a large body of work in complex network analysis has focused on developing generative models of graphs which mimick real-wor...
In this paper we present three algorithms that build graph layouts for undirected, weighted graphs. Our goal is to generate layouts that are consistent with the weights in the gra...
Feature selection is an important task in order to achieve better generalizability in high dimensional learning, and structure learning of Markov random fields (MRFs) can automat...