Scene recognition with bag of visual nouns and prepositions

John Stalbaum, Hee Won Chae, Jae-Bok Song

Research output: Contribution to journalArticle

Abstract

The loop closure problem is central to topological simultaneous localization and mapping (SLAM); by associating features between distant portions of a trajectory, the odometry error that has accumulated between two observations can be eliminated and a more consistent map can be built. Bayesian pattern recognition techniques such as bag of visual words (BoVW) have recently shown outstanding results in solving the loop closure problem completely in image space using very simple, inexpensive cameras, without the requirement for highly accurate metric information, 3D reconstruction, or camera calibration. In this paper, a modified BoVW descriptor that incorporates simple geometric relationships within an image is used with the fast appearance-based mapping (FAB-MAP) algorithm. In direct comparisons with the traditional BoVW descriptor, an improved recall rate is observed with an acceptable increase in computational time. The proposal of a BoVW-compatible descriptor and the use of the proposed descriptor with a well-known BoVW classifier demonstrate the ability of the BoVW metaphor to be generalized, which could pave the way for more various BoVW descriptors in the same way that many individual visual feature descriptors exist within the computer vision community.

Original languageEnglish
Pages (from-to)115-125
Number of pages11
JournalIntelligent Service Robotics
Volume8
Issue number2
DOIs
Publication statusPublished - 2015 Apr 1

Fingerprint

Cameras
Computer vision
Pattern recognition
Classifiers
Trajectories
Calibration

Keywords

  • Bag of visual words
  • Loop closure
  • Place recognition
  • Scene recognition
  • SLAM

ASJC Scopus subject areas

  • Mechanical Engineering
  • Engineering (miscellaneous)
  • Artificial Intelligence
  • Computational Mechanics

Cite this

Scene recognition with bag of visual nouns and prepositions. / Stalbaum, John; Chae, Hee Won; Song, Jae-Bok.

In: Intelligent Service Robotics, Vol. 8, No. 2, 01.04.2015, p. 115-125.

Research output: Contribution to journalArticle

Stalbaum, John ; Chae, Hee Won ; Song, Jae-Bok. / Scene recognition with bag of visual nouns and prepositions. In: Intelligent Service Robotics. 2015 ; Vol. 8, No. 2. pp. 115-125.
@article{311cebd8d0be404c8e5eb9b34afc423c,
title = "Scene recognition with bag of visual nouns and prepositions",
abstract = "The loop closure problem is central to topological simultaneous localization and mapping (SLAM); by associating features between distant portions of a trajectory, the odometry error that has accumulated between two observations can be eliminated and a more consistent map can be built. Bayesian pattern recognition techniques such as bag of visual words (BoVW) have recently shown outstanding results in solving the loop closure problem completely in image space using very simple, inexpensive cameras, without the requirement for highly accurate metric information, 3D reconstruction, or camera calibration. In this paper, a modified BoVW descriptor that incorporates simple geometric relationships within an image is used with the fast appearance-based mapping (FAB-MAP) algorithm. In direct comparisons with the traditional BoVW descriptor, an improved recall rate is observed with an acceptable increase in computational time. The proposal of a BoVW-compatible descriptor and the use of the proposed descriptor with a well-known BoVW classifier demonstrate the ability of the BoVW metaphor to be generalized, which could pave the way for more various BoVW descriptors in the same way that many individual visual feature descriptors exist within the computer vision community.",
keywords = "Bag of visual words, Loop closure, Place recognition, Scene recognition, SLAM",
author = "John Stalbaum and Chae, {Hee Won} and Jae-Bok Song",
year = "2015",
month = "4",
day = "1",
doi = "10.1007/s11370-015-0167-0",
language = "English",
volume = "8",
pages = "115--125",
journal = "Intelligent Service Robotics",
issn = "1861-2776",
publisher = "Springer Verlag",
number = "2",

}

TY - JOUR

T1 - Scene recognition with bag of visual nouns and prepositions

AU - Stalbaum, John

AU - Chae, Hee Won

AU - Song, Jae-Bok

PY - 2015/4/1

Y1 - 2015/4/1

N2 - The loop closure problem is central to topological simultaneous localization and mapping (SLAM); by associating features between distant portions of a trajectory, the odometry error that has accumulated between two observations can be eliminated and a more consistent map can be built. Bayesian pattern recognition techniques such as bag of visual words (BoVW) have recently shown outstanding results in solving the loop closure problem completely in image space using very simple, inexpensive cameras, without the requirement for highly accurate metric information, 3D reconstruction, or camera calibration. In this paper, a modified BoVW descriptor that incorporates simple geometric relationships within an image is used with the fast appearance-based mapping (FAB-MAP) algorithm. In direct comparisons with the traditional BoVW descriptor, an improved recall rate is observed with an acceptable increase in computational time. The proposal of a BoVW-compatible descriptor and the use of the proposed descriptor with a well-known BoVW classifier demonstrate the ability of the BoVW metaphor to be generalized, which could pave the way for more various BoVW descriptors in the same way that many individual visual feature descriptors exist within the computer vision community.

AB - The loop closure problem is central to topological simultaneous localization and mapping (SLAM); by associating features between distant portions of a trajectory, the odometry error that has accumulated between two observations can be eliminated and a more consistent map can be built. Bayesian pattern recognition techniques such as bag of visual words (BoVW) have recently shown outstanding results in solving the loop closure problem completely in image space using very simple, inexpensive cameras, without the requirement for highly accurate metric information, 3D reconstruction, or camera calibration. In this paper, a modified BoVW descriptor that incorporates simple geometric relationships within an image is used with the fast appearance-based mapping (FAB-MAP) algorithm. In direct comparisons with the traditional BoVW descriptor, an improved recall rate is observed with an acceptable increase in computational time. The proposal of a BoVW-compatible descriptor and the use of the proposed descriptor with a well-known BoVW classifier demonstrate the ability of the BoVW metaphor to be generalized, which could pave the way for more various BoVW descriptors in the same way that many individual visual feature descriptors exist within the computer vision community.

KW - Bag of visual words

KW - Loop closure

KW - Place recognition

KW - Scene recognition

KW - SLAM

UR - http://www.scopus.com/inward/record.url?scp=84939997657&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84939997657&partnerID=8YFLogxK

U2 - 10.1007/s11370-015-0167-0

DO - 10.1007/s11370-015-0167-0

M3 - Article

AN - SCOPUS:84939997657

VL - 8

SP - 115

EP - 125

JO - Intelligent Service Robotics

JF - Intelligent Service Robotics

SN - 1861-2776

IS - 2

ER -