Traditional Markov network structure learning algorithms perform a search for globally useful features. However, these algorithms are often slow and prone to finding local optima d...
— Learning Vector Quantization (LVQ) is a popular class of nearest prototype classifiers for multiclass classification. Learning algorithms from this family are widely used becau...
3D object detection and importance regression/ranking are at the core for semantically interpreting 3D medical images of computer aided diagnosis (CAD). In this paper, we propose ...
Le Lu, Jinbo Bi, Matthias Wolf, Marcos Salganicoff
1 The techniques for image analysis and classi cation generally consider the image sample labels xed and without uncertainties. The rank regression problem is studied in this pape...
The two main plagues of image restoration are oscillations and smoothing. Traditional image restoration techniques prevent parasitic oscillations by resorting to smooth regulariza...