How the Number of Interest Points Affect Scene Classification

Wenjie XIE; De XU; Shuoyan LIU; Yingjun TANG

doi:10.1587/transinf.E93.D.930

How the Number of Interest Points Affect Scene Classification

Wenjie XIE, De XU, Shuoyan LIU, Yingjun TANG

Full Text Views

0

Cite this

Summary :

This paper focuses on the relationship between the number of interest points and the accuracy rate in scene classification. Here, we accept the common belief that more interest points can generate higher accuracy. But, few effort have been done in this field. In order to validate this viewpoint, in our paper, extensive experiments based on bag of words method are implemented. In particular, three different SIFT descriptors and five feature selection methods are adopted to change the number of interest points. As innovation point, we propose a novel dense SIFT descriptor named Octave Dense SIFT, which can generate more interest points and higher accuracy, and a new feature selection method called number mutual information (NMI), which has better robustness than other feature selection methods. Experimental results show that the number of interest points can aggressively affect classification accuracy.

Publication: IEICE TRANSACTIONS on Information Vol.E93-D No.4 pp.930-933

Publication Date: 2010/04/01

Publicized

Online ISSN: 1745-1361

DOI: 10.1587/transinf.E93.D.930

Type of Manuscript: LETTER

Category: Image Recognition, Computer Vision

Cite this

Copy

Wenjie XIE, De XU, Shuoyan LIU, Yingjun TANG, "How the Number of Interest Points Affect Scene Classification" in IEICE TRANSACTIONS on Information, vol. E93-D, no. 4, pp. 930-933, April 2010, doi: 10.1587/transinf.E93.D.930.
Abstract: This paper focuses on the relationship between the number of interest points and the accuracy rate in scene classification. Here, we accept the common belief that more interest points can generate higher accuracy. But, few effort have been done in this field. In order to validate this viewpoint, in our paper, extensive experiments based on bag of words method are implemented. In particular, three different SIFT descriptors and five feature selection methods are adopted to change the number of interest points. As innovation point, we propose a novel dense SIFT descriptor named Octave Dense SIFT, which can generate more interest points and higher accuracy, and a new feature selection method called number mutual information (NMI), which has better robustness than other feature selection methods. Experimental results show that the number of interest points can aggressively affect classification accuracy.
URL: https://global.ieice.org/en_transactions/information/10.1587/transinf.E93.D.930/_p

Copy

@ARTICLE{e93-d_4_930,
author={Wenjie XIE, De XU, Shuoyan LIU, Yingjun TANG, },
journal={IEICE TRANSACTIONS on Information},
title={How the Number of Interest Points Affect Scene Classification},
year={2010},
volume={E93-D},
number={4},
pages={930-933},
abstract={This paper focuses on the relationship between the number of interest points and the accuracy rate in scene classification. Here, we accept the common belief that more interest points can generate higher accuracy. But, few effort have been done in this field. In order to validate this viewpoint, in our paper, extensive experiments based on bag of words method are implemented. In particular, three different SIFT descriptors and five feature selection methods are adopted to change the number of interest points. As innovation point, we propose a novel dense SIFT descriptor named Octave Dense SIFT, which can generate more interest points and higher accuracy, and a new feature selection method called number mutual information (NMI), which has better robustness than other feature selection methods. Experimental results show that the number of interest points can aggressively affect classification accuracy.},
keywords={},
doi={10.1587/transinf.E93.D.930},
ISSN={1745-1361},
month={April},}

Copy

TY - JOUR
TI - How the Number of Interest Points Affect Scene Classification
T2 - IEICE TRANSACTIONS on Information
SP - 930
EP - 933
AU - Wenjie XIE
AU - De XU
AU - Shuoyan LIU
AU - Yingjun TANG
PY - 2010
DO - 10.1587/transinf.E93.D.930
JO - IEICE TRANSACTIONS on Information
SN - 1745-1361
VL - E93-D
IS - 4
JA - IEICE TRANSACTIONS on Information
Y1 - April 2010
AB - This paper focuses on the relationship between the number of interest points and the accuracy rate in scene classification. Here, we accept the common belief that more interest points can generate higher accuracy. But, few effort have been done in this field. In order to validate this viewpoint, in our paper, extensive experiments based on bag of words method are implemented. In particular, three different SIFT descriptors and five feature selection methods are adopted to change the number of interest points. As innovation point, we propose a novel dense SIFT descriptor named Octave Dense SIFT, which can generate more interest points and higher accuracy, and a new feature selection method called number mutual information (NMI), which has better robustness than other feature selection methods. Experimental results show that the number of interest points can aggressively affect classification accuracy.
ER -