A new feature descriptor is presented for object and scene recognition. The new approach, called CDIKP, uniquely combines the scale-invariant feature detection with a robust proje...
This paper presents an exploratory study of ten feature location techniques that use various combinations of textual, dynamic, and static analyses. Unlike previous studies, the ap...
Abstract. Lexical variance in biomedical texts poses a challenge to automatic protein relation mining. We therefore propose a new approach that relies only on more general language...
Timur Fayruzov, Martine De Cock, Chris Cornelis, V...
Abstract. We present a method that improves the results of network intrusion detection by integration of several anomaly detection algorithms through trust and reputation models. O...
Abstract. This paper describes two procedures for generating very short summaries for documents from the DUC-2003 competition: a chunk extraction method based on syntactic dependen...