In this paper, a novel method for reducing the runtime complexity of a support vector machine classifier is presented. The new training algorithm is fast and simple. This is achiev...
In this paper, we present a method for binary image comparison. For binary images, intensity information is poor and shape extraction is often difficult. Therefore binary images h...
We study online classification of isolated handwritten symbols using distance measures on spaces of curves. We compare three distance-based measures on a vector space representatio...
Data clustering is a difficult problem due to the complex and heterogeneous natures of multidimensional data. To improve clustering accuracy, we propose a scheme to capture the lo...
Abstract--This paper presents a new region-based unified tensor level set model for image segmentation. This model introduces a three-order tensor to comprehensively depict feature...