To access data sources on the Web, a crucial step is wrapping, which translates query responses, rendered in textual HTML, back into their relational form. Traditionally, this pro...
Shui-Lung Chuang, Kevin Chen-Chuan Chang, ChengXia...
The output of boolean association rule mining algorithms is often too large for manual examination. For dense datasets, it is often impractical to even generate all frequent items...
We present the Higher Order Proxy Neighborhoods (HOPS) approach to modeling higher order neighborhoods in Markov Random Fields (MRFs). HOPS incorporates more context information i...
Abstract. This paper addresses the problem of automatically segmenting bone structures in low resolution clinical MRI datasets. The novel aspect of the proposed method is the combi...
Probabilistic models are extensively used in medical image segmentation. Most of them employ parametric representations of densities and make idealizing assumptions, e.g. normal di...