We introduce a new approach to image reconstruction from highly incomplete data. The available data are assumed to be a small collection of spectral coef?cients of an arbitrary li...
Karen O. Egiazarian, Alessandro Foi, Vladimir Katk...
Information diffusion, viral marketing, and collective classification all attempt to model and exploit the relationships in a network to make inferences about the labels of nodes....
Visual data comprises of multi-scale and inhomogeneous signals. In this paper, we exploit these characteristics and develop an adaptive data approximation technique based on a hie...
Abstract. This paper proposes to exploit content and usage information to rearrange an inverted index for a full-text IR system. The idea is to merge the entries of two frequently ...
Large search engines process thousands of queries per second on billions of pages, making query processing a major factor in their operating costs. This has led to a lot of resear...