Abstract. The Mumford-Shah model has been one of the most powerful models in image segmentation and denoising. The optimization of the multiphase Mumford-Shah energy functional has...
Abstract—We present an algorithm for removing timefrequency components, found by a standard Gabor transform, of a “real-world” sound while causing no audible difference to th...
Background: Several supervised and unsupervised learning tools are available to classify functional genomics data. However, relatively less attention has been given to exploratory...
The nDCG measure has proven to be a popular measure of retrieval effectiveness utilizing graded relevance judgments. However, a number of different instantiations of nDCG exist, d...
Virtual evidence (VE), first introduced by (Pearl, 1988), provides a convenient way of incorporating prior knowledge into Bayesian networks. This work generalizes the use of VE to...