A Novel Graph Embedding Framework for Object Recognition
Abstract
A great deal of research works have been devoted to understand image contents. In this field many well-known methods exploit Bag of Words ( BoW ) features describing image contents as appearance frequency histogram of visual words. These approaches have a main drawback, the location information and the relationships between features are lost. To overcame this limitation we propose a novel methodology for the Object recognition task. A digital image is described as a feature vector computed by means of a new graph embedding paradigm on the Attributed Relational SIFT Regions Graph. The final classification is performed by using Logistic Label Propagation classifier. Our framework is evaluated on standard databases (such as ETH- $80$ , COIL- $100$ and ALOI) and the achieved results compared with those obtained by well-known methodologies confirm its quality.
Cite
Text
Manzo et al. "A Novel Graph Embedding Framework for Object Recognition." European Conference on Computer Vision, 2014. doi:10.1007/978-3-319-16220-1_24Markdown
[Manzo et al. "A Novel Graph Embedding Framework for Object Recognition." European Conference on Computer Vision, 2014.](https://mlanthology.org/eccv/2014/manzo2014eccv-novel/) doi:10.1007/978-3-319-16220-1_24BibTeX
@inproceedings{manzo2014eccv-novel,
title = {{A Novel Graph Embedding Framework for Object Recognition}},
author = {Manzo, Mario and Pellino, Simone and Petrosino, Alfredo and Rozza, Alessandro},
booktitle = {European Conference on Computer Vision},
year = {2014},
pages = {341-352},
doi = {10.1007/978-3-319-16220-1_24},
url = {https://mlanthology.org/eccv/2014/manzo2014eccv-novel/}
}