An efficient incremental face annotation for large scale web services

Kwontaeg Choi, Kar Ann Toh, Hyeran Byun

Research output: Contribution to journalArticle

5 Citations (Scopus)

Abstract

This paper proposes an incremental face annotation framework for sharing and publishing photographs which contain faces under a large scale web platform such as a social network service with millions of users. Unlike the conventional face recognition environment addressed by most existing works, the image databases being accessed by the large pool of users can be huge and frequently updated. A reasonable way to efficiently annotate such huge databases is to accommodate an adaptation of model parameters without the need to retrain the model all over again when new data arrives. In this work, we are particularly interested in the following issues related to increment of data: (i) the huge number of images being added at each instant, (ii) the large number of users joining the web each day, and (iii) the large number of classification systems being added at each period. We propose an efficient recursive estimation method to handle these data increment issues. Our experiments on several databases show that our proposed method achieves an almost constant execution time with comparable accuracy relative to those state-of-the-art incremental versions of principal component analysis, linear discriminant analysis and support vector machine.

Original languageEnglish
Pages (from-to)197-214
Number of pages18
JournalTelecommunication Systems
Volume47
Issue number3-4
DOIs
Publication statusPublished - 2011 Aug 1

Fingerprint

Web services
Discriminant analysis
Face recognition
Joining
Principal component analysis
Support vector machines
Experiments

All Science Journal Classification (ASJC) codes

  • Electrical and Electronic Engineering

Cite this

@article{b3fed94b1fa6496c8818b21822cea428,
title = "An efficient incremental face annotation for large scale web services",
abstract = "This paper proposes an incremental face annotation framework for sharing and publishing photographs which contain faces under a large scale web platform such as a social network service with millions of users. Unlike the conventional face recognition environment addressed by most existing works, the image databases being accessed by the large pool of users can be huge and frequently updated. A reasonable way to efficiently annotate such huge databases is to accommodate an adaptation of model parameters without the need to retrain the model all over again when new data arrives. In this work, we are particularly interested in the following issues related to increment of data: (i) the huge number of images being added at each instant, (ii) the large number of users joining the web each day, and (iii) the large number of classification systems being added at each period. We propose an efficient recursive estimation method to handle these data increment issues. Our experiments on several databases show that our proposed method achieves an almost constant execution time with comparable accuracy relative to those state-of-the-art incremental versions of principal component analysis, linear discriminant analysis and support vector machine.",
author = "Kwontaeg Choi and Toh, {Kar Ann} and Hyeran Byun",
year = "2011",
month = "8",
day = "1",
doi = "10.1007/s11235-010-9312-4",
language = "English",
volume = "47",
pages = "197--214",
journal = "Telecommunication Systems",
issn = "1018-4864",
publisher = "Springer Netherlands",
number = "3-4",

}

An efficient incremental face annotation for large scale web services. / Choi, Kwontaeg; Toh, Kar Ann; Byun, Hyeran.

In: Telecommunication Systems, Vol. 47, No. 3-4, 01.08.2011, p. 197-214.

Research output: Contribution to journalArticle

TY - JOUR

T1 - An efficient incremental face annotation for large scale web services

AU - Choi, Kwontaeg

AU - Toh, Kar Ann

AU - Byun, Hyeran

PY - 2011/8/1

Y1 - 2011/8/1

N2 - This paper proposes an incremental face annotation framework for sharing and publishing photographs which contain faces under a large scale web platform such as a social network service with millions of users. Unlike the conventional face recognition environment addressed by most existing works, the image databases being accessed by the large pool of users can be huge and frequently updated. A reasonable way to efficiently annotate such huge databases is to accommodate an adaptation of model parameters without the need to retrain the model all over again when new data arrives. In this work, we are particularly interested in the following issues related to increment of data: (i) the huge number of images being added at each instant, (ii) the large number of users joining the web each day, and (iii) the large number of classification systems being added at each period. We propose an efficient recursive estimation method to handle these data increment issues. Our experiments on several databases show that our proposed method achieves an almost constant execution time with comparable accuracy relative to those state-of-the-art incremental versions of principal component analysis, linear discriminant analysis and support vector machine.

AB - This paper proposes an incremental face annotation framework for sharing and publishing photographs which contain faces under a large scale web platform such as a social network service with millions of users. Unlike the conventional face recognition environment addressed by most existing works, the image databases being accessed by the large pool of users can be huge and frequently updated. A reasonable way to efficiently annotate such huge databases is to accommodate an adaptation of model parameters without the need to retrain the model all over again when new data arrives. In this work, we are particularly interested in the following issues related to increment of data: (i) the huge number of images being added at each instant, (ii) the large number of users joining the web each day, and (iii) the large number of classification systems being added at each period. We propose an efficient recursive estimation method to handle these data increment issues. Our experiments on several databases show that our proposed method achieves an almost constant execution time with comparable accuracy relative to those state-of-the-art incremental versions of principal component analysis, linear discriminant analysis and support vector machine.

UR - http://www.scopus.com/inward/record.url?scp=79958135963&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=79958135963&partnerID=8YFLogxK

U2 - 10.1007/s11235-010-9312-4

DO - 10.1007/s11235-010-9312-4

M3 - Article

VL - 47

SP - 197

EP - 214

JO - Telecommunication Systems

JF - Telecommunication Systems

SN - 1018-4864

IS - 3-4

ER -