Improving classification accuracy of image categories using local descriptors with supplemental information

Kazuya Ueki, Youhei Shiraishi, Naohiro Tawara, Tetsunori Kobayashi

Research output: Contribution to journalArticle

Abstract

In this paper we address the problem of image classification by embedding the spatial information into the local descriptor. In our method, we directly concatenate (x,y) coordinates of an image into the original feature vector. This simple idea can perform well in the object category classification even though the feature vector size is almost the same as the conventional approach. Results are reported for classification of the Caltech-101 dataset and our methods are found to produce consistently better results compared with traditional Bag-of-Features approaches in all experiments.

Original languageEnglish
Pages (from-to)1144-1149
Number of pages6
JournalSeimitsu Kogaku Kaishi/Journal of the Japan Society for Precision Engineering
Volume80
Issue number12
DOIs
Publication statusPublished - 2014 Dec 1
Externally publishedYes

Fingerprint

Image classification
Experiments

Keywords

  • Bag-of-features
  • Generic object recognition
  • Image category classification
  • Local features
  • Spacial information

ASJC Scopus subject areas

  • Mechanical Engineering

Cite this

Improving classification accuracy of image categories using local descriptors with supplemental information. / Ueki, Kazuya; Shiraishi, Youhei; Tawara, Naohiro; Kobayashi, Tetsunori.

In: Seimitsu Kogaku Kaishi/Journal of the Japan Society for Precision Engineering, Vol. 80, No. 12, 01.12.2014, p. 1144-1149.

Research output: Contribution to journalArticle

@article{27e692c999d148cda1f195a343e516ef,
title = "Improving classification accuracy of image categories using local descriptors with supplemental information",
abstract = "In this paper we address the problem of image classification by embedding the spatial information into the local descriptor. In our method, we directly concatenate (x,y) coordinates of an image into the original feature vector. This simple idea can perform well in the object category classification even though the feature vector size is almost the same as the conventional approach. Results are reported for classification of the Caltech-101 dataset and our methods are found to produce consistently better results compared with traditional Bag-of-Features approaches in all experiments.",
keywords = "Bag-of-features, Generic object recognition, Image category classification, Local features, Spacial information",
author = "Kazuya Ueki and Youhei Shiraishi and Naohiro Tawara and Tetsunori Kobayashi",
year = "2014",
month = "12",
day = "1",
doi = "10.2493/jjspe.80.1144",
language = "English",
volume = "80",
pages = "1144--1149",
journal = "Seimitsu Kogaku Kaishi/Journal of the Japan Society for Precision Engineering",
issn = "0912-0289",
publisher = "Japan Society for Precision Engineering",
number = "12",

}

TY - JOUR

T1 - Improving classification accuracy of image categories using local descriptors with supplemental information

AU - Ueki, Kazuya

AU - Shiraishi, Youhei

AU - Tawara, Naohiro

AU - Kobayashi, Tetsunori

PY - 2014/12/1

Y1 - 2014/12/1

N2 - In this paper we address the problem of image classification by embedding the spatial information into the local descriptor. In our method, we directly concatenate (x,y) coordinates of an image into the original feature vector. This simple idea can perform well in the object category classification even though the feature vector size is almost the same as the conventional approach. Results are reported for classification of the Caltech-101 dataset and our methods are found to produce consistently better results compared with traditional Bag-of-Features approaches in all experiments.

AB - In this paper we address the problem of image classification by embedding the spatial information into the local descriptor. In our method, we directly concatenate (x,y) coordinates of an image into the original feature vector. This simple idea can perform well in the object category classification even though the feature vector size is almost the same as the conventional approach. Results are reported for classification of the Caltech-101 dataset and our methods are found to produce consistently better results compared with traditional Bag-of-Features approaches in all experiments.

KW - Bag-of-features

KW - Generic object recognition

KW - Image category classification

KW - Local features

KW - Spacial information

UR - http://www.scopus.com/inward/record.url?scp=84988962630&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84988962630&partnerID=8YFLogxK

U2 - 10.2493/jjspe.80.1144

DO - 10.2493/jjspe.80.1144

M3 - Article

VL - 80

SP - 1144

EP - 1149

JO - Seimitsu Kogaku Kaishi/Journal of the Japan Society for Precision Engineering

JF - Seimitsu Kogaku Kaishi/Journal of the Japan Society for Precision Engineering

SN - 0912-0289

IS - 12

ER -