Domain-Size Pooling in Local Descriptors: DSP-SIFT

Jingming Dong, Stefano Soatto; The IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2015, pp. 5097-5106

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.

Related Material


[pdf]
[bibtex]
@InProceedings{Dong_2015_CVPR,
author = {Dong, Jingming and Soatto, Stefano},
title = {Domain-Size Pooling in Local Descriptors: DSP-SIFT},
booktitle = {The IEEE Conference on Computer Vision and Pattern Recognition (CVPR)},
month = {June},
year = {2015}
}