Abstract- The large majority of existing clustering algorithms are centered around the notion of a feature, that is, individual data items are represented by their intrinsic proper...
In earlier papers, it was shown that recognizing persons using their brain patterns evoked during visual stimulus is possible. In this paper, several modifications are proposed to...
We report an automatic feature discovery method that achieves results comparable to a manually chosen, larger feature set on a document image content extraction problem: the locat...
Learning useful and predictable features from past workloads and exploiting them well is a major source of improvement in many operating system problems. We review known parallel ...
We propose a geometric approach to 3D motion segmentation from point correspondences in three perspective views. We demonstrate that after applying a polynomial embedding to the po...