We consider the problem of computing all-pair correlations in a warehouse containing a large number (e.g., tens of thousands) of time-series (or, signals). The problem arises in a...
"Pattern recognition techniques are concerned with the theory and algorithms of putting abstract objects, e.g., measurements made on physical objects, into categories. Typical...
Stable local feature detection and representation is a fundamental component of many image registration and object recognition algorithms. Mikolajczyk and Schmid [14] recently eva...
A method that combines shape-based object recognition and image segmentation is proposed for shape retrieval from images. Given a shape prior represented in a multiscale curvature...
This paper presents a new incremental learning solution for Linear Discriminant Analysis (LDA). We apply the concept of the sufficient spanning set approximation in each update st...