We present a novel clustering method using HMM parameter space and eigenvector decomposition. Unlike the existing methods, our algorithm can cluster both constant and variable leng...
The scalability problem in data mining involves the development of methods for handling large databases with limited computational resources. In this paper, we present a two-phase...
This paper casts the problem of point-set alignment via Procrustes analysis into a maximum likelihood framework using the EM algorithm. The aim is to improve the robustness of the...
Multiple sensors are commonly fused to improve the detection and recognition performance of computer vision and pattern recognition systems. The traditional approach to determine ...