Synthesis of high-resolution facial image based on top-down learning

Bon W. Hwang, Jeong Seon Park, Seong Whan Lee

Research output: Contribution to journalArticle

2 Citations (Scopus)

Abstract

This paper proposes a method of synthesizing a high-resolution facial image from a low-resolution facial image based on top-down learning. A face is represented by a linear combination of prototypes of shape and texture. With the shape and texture information about the pixels in an given low-resolution facial image, we can estimate optimal coefficients for a linear combination of prototypes of shape and those of texture by solving least square minimization. Then high-resolution facial image can be synthesized by using the optimal coefficients for linear combination of the high-resolution prototypes. The encouraging results of the proposed method show that our method can be used to increase the performance of the face recognition by applying our method to enhance the low-resolution facial images captured at surveillance systems.

Original languageEnglish
Pages (from-to)377-384
Number of pages8
JournalLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume2688
Publication statusPublished - 2003 Dec 1

Fingerprint

High Resolution
Textures
Learning
Synthesis
Linear Combination
Texture
Prototype
Face recognition
Least-Squares Analysis
Pixels
Coefficient
Face Recognition
Surveillance
Least Squares
Pixel
Face
Estimate

ASJC Scopus subject areas

  • Biochemistry, Genetics and Molecular Biology(all)
  • Computer Science(all)
  • Theoretical Computer Science

Cite this

@article{19c43f2b640b4b8182d5ee53586cdcf7,
title = "Synthesis of high-resolution facial image based on top-down learning",
abstract = "This paper proposes a method of synthesizing a high-resolution facial image from a low-resolution facial image based on top-down learning. A face is represented by a linear combination of prototypes of shape and texture. With the shape and texture information about the pixels in an given low-resolution facial image, we can estimate optimal coefficients for a linear combination of prototypes of shape and those of texture by solving least square minimization. Then high-resolution facial image can be synthesized by using the optimal coefficients for linear combination of the high-resolution prototypes. The encouraging results of the proposed method show that our method can be used to increase the performance of the face recognition by applying our method to enhance the low-resolution facial images captured at surveillance systems.",
author = "Hwang, {Bon W.} and Park, {Jeong Seon} and Lee, {Seong Whan}",
year = "2003",
month = "12",
day = "1",
language = "English",
volume = "2688",
pages = "377--384",
journal = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
issn = "0302-9743",
publisher = "Springer Verlag",

}

TY - JOUR

T1 - Synthesis of high-resolution facial image based on top-down learning

AU - Hwang, Bon W.

AU - Park, Jeong Seon

AU - Lee, Seong Whan

PY - 2003/12/1

Y1 - 2003/12/1

N2 - This paper proposes a method of synthesizing a high-resolution facial image from a low-resolution facial image based on top-down learning. A face is represented by a linear combination of prototypes of shape and texture. With the shape and texture information about the pixels in an given low-resolution facial image, we can estimate optimal coefficients for a linear combination of prototypes of shape and those of texture by solving least square minimization. Then high-resolution facial image can be synthesized by using the optimal coefficients for linear combination of the high-resolution prototypes. The encouraging results of the proposed method show that our method can be used to increase the performance of the face recognition by applying our method to enhance the low-resolution facial images captured at surveillance systems.

AB - This paper proposes a method of synthesizing a high-resolution facial image from a low-resolution facial image based on top-down learning. A face is represented by a linear combination of prototypes of shape and texture. With the shape and texture information about the pixels in an given low-resolution facial image, we can estimate optimal coefficients for a linear combination of prototypes of shape and those of texture by solving least square minimization. Then high-resolution facial image can be synthesized by using the optimal coefficients for linear combination of the high-resolution prototypes. The encouraging results of the proposed method show that our method can be used to increase the performance of the face recognition by applying our method to enhance the low-resolution facial images captured at surveillance systems.

UR - http://www.scopus.com/inward/record.url?scp=21144440814&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=21144440814&partnerID=8YFLogxK

M3 - Article

VL - 2688

SP - 377

EP - 384

JO - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

JF - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SN - 0302-9743

ER -