Domain-Size Pooling in Local Descriptors: DSP-SIFT

CVPR 2015

doi:10.1109/CVPR.2015.7299145 /cvpr/2015/dong2015cvpr-domainsize/

Abstract

We introduce a simple modification of local image descriptors, such as SIFT, based on pooling gradient orientations across different domain sizes, in addition to spatial locations. The resulting descriptor, which we call DSP-SIFT, outperforms other methods in wide-baseline matching benchmarks, including those based on convolutional neural networks, despite having the same dimension of SIFT and requiring no training.

PDF CVPR Semantic Scholar

Cite

Text

Dong and Soatto. "Domain-Size Pooling in Local Descriptors: DSP-SIFT." Conference on Computer Vision and Pattern Recognition, 2015. doi:10.1109/CVPR.2015.7299145

Markdown

[Dong and Soatto. "Domain-Size Pooling in Local Descriptors: DSP-SIFT." Conference on Computer Vision and Pattern Recognition, 2015.](https://mlanthology.org/cvpr/2015/dong2015cvpr-domainsize/) doi:10.1109/CVPR.2015.7299145

BibTeX

@inproceedings{dong2015cvpr-domainsize,
  title     = {{Domain-Size Pooling in Local Descriptors: DSP-SIFT}},
  author    = {Dong, Jingming and Soatto, Stefano},
  booktitle = {Conference on Computer Vision and Pattern Recognition},
  year      = {2015},
  doi       = {10.1109/CVPR.2015.7299145},
  url       = {https://mlanthology.org/cvpr/2015/dong2015cvpr-domainsize/}
}