BibTeX
@MISC{Mahdaviani05fastobject,
author = {Maryam Mahdaviani},
title = {Fast Object Class Recognition},
year = {2005}
}
Share
 |
 |
 |
 |
||
OpenURL
Abstract
We propose a novel approach for object class recognition using scale invariant features and Gaussian Processes as our kernel-based classifier. We measure the performance of this approach in two stages: predicting the presence of a class of objects in images and localizing them. Our object class recognition method is comparable to other state-of-the-art approaches. Furthermore, we propose sophisticated numerical methods for reducing the computational cost of Gaussian Processes. Speed-ups are achieved using Krylov Subspace and Dual-Tree methods. We also reduced the storage requirement from O(M 2) to O(M). We trained and tested our algorithms on the PASCAL collection of images, which is a common database in this research area. ii Contents Abstract ii
Keyphrases
fast object class recognition gaussian process common database pascal collection krylov subspace scale invariant feature sophisticated numerical method state-of-the-art approach kernel-based classifier object class recognition ii content abstract ii novel approach dual-tree method computational cost storage requirement research area object class recognition method
