Selection of Universal Features for Image Classification
Abstract
Neuromimetic algorithms, such as the HMAX algorithm, have been very successful in image classification tasks. However, current implementations of these algorithms do not scale well to large datasets. Often, target-specific features or patches are "learned" ahead of time and then correlated with test images during feature extraction. In this paper, we develop a novel method for selecting a single set of universal features that enables classification across a broad range of image classes. Our method trains multiple Random Forest classifiers using a large dictionary of features and then combines them using a majority voting scheme. This enables the selection of the most discriminative patches based on feature importance measures. Experiments demonstrate the viability of this method using HMAX features as well as the tradeoff between the number of universal features, classification performance, and processing time.
Cite
Text
Rodriguez et al. "Selection of Universal Features for Image Classification." IEEE/CVF Winter Conference on Applications of Computer Vision, 2014. doi:10.1109/WACV.2014.6836078Markdown
[Rodriguez et al. "Selection of Universal Features for Image Classification." IEEE/CVF Winter Conference on Applications of Computer Vision, 2014.](https://mlanthology.org/wacv/2014/rodriguez2014wacv-selection/) doi:10.1109/WACV.2014.6836078BibTeX
@inproceedings{rodriguez2014wacv-selection,
title = {{Selection of Universal Features for Image Classification}},
author = {Rodriguez, Pedro A. and Drenkow, Nathan and DeMenthon, Daniel and Koterba, Zachary H. and Kauffman, Kathleen and Cornish, Duane and Paulhamus, Bart L. and Vogelstein, R. Jacob},
booktitle = {IEEE/CVF Winter Conference on Applications of Computer Vision},
year = {2014},
pages = {355-362},
doi = {10.1109/WACV.2014.6836078},
url = {https://mlanthology.org/wacv/2014/rodriguez2014wacv-selection/}
}