Vast amounts of text on the Web are unstructured and ungrammatical, such as classified ads, auction listings, forum postings, etc. We call such text “posts.” Despite their in...
This paper presents a method for unsupervised discovery of semantic patterns. Semantic patterns are useful for a variety of text understanding tasks, in particular for locating ev...
Abstract. We propose a novel method for unsupervised class segmentation on a set of images. It alternates between segmenting object instances and learning a class model. The method...
The photorealistic modeling of large-scale objects, such as urban scenes, requires the combination of range sensing technology and digital photography. In this paper, we attack th...
In this paper, we present an accurate and realtime PE-Miner framework that automatically extracts distinguishing features from portable executables (PE) to detect zero-day (i.e. pr...
M. Zubair Shafiq, S. Momina Tabish, Fauzan Mirza, ...