Converting facial expressions using recognition-based analysis of image sequences

Takahiro Otsuka, Jun Ohya

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

A method for converting one person’s facial expression into another person’s is proposed. The sequence of the feature vector for each expression is modeled by using HMM with the hidden states corresponding to the different muscle conditions (relaxed, contracting, and the end of contraction). The probability of each state is evaluated for each frame and the contraction rate of each muscle is obtained from the probability of each state using a matrix representing the characteristics of other people’s expressions. The experiments showed the superior realism of the expression generated by our proposed method.

Original languageEnglish
Title of host publicationLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
PublisherSpringer Verlag
Pages703-710
Number of pages8
Volume1352
ISBN (Print)3540639314, 9783540639312
Publication statusPublished - 1997
Externally publishedYes
Event3rd Asian Conference on Computer Vision, ACCV 1998 - Hong Kong, Hong Kong
Duration: 1998 Jan 81998 Jan 10

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume1352
ISSN (Print)03029743
ISSN (Electronic)16113349

Other

Other3rd Asian Conference on Computer Vision, ACCV 1998
CountryHong Kong
CityHong Kong
Period98/1/898/1/10

Fingerprint

Facial Expression Recognition
Image Sequence
Muscle
Contraction
Person
Facial Expression
Feature Vector
Experiments
Experiment

ASJC Scopus subject areas

  • Computer Science(all)
  • Theoretical Computer Science

Cite this

Otsuka, T., & Ohya, J. (1997). Converting facial expressions using recognition-based analysis of image sequences. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 1352, pp. 703-710). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 1352). Springer Verlag.

Converting facial expressions using recognition-based analysis of image sequences. / Otsuka, Takahiro; Ohya, Jun.

Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). Vol. 1352 Springer Verlag, 1997. p. 703-710 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 1352).

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Otsuka, T & Ohya, J 1997, Converting facial expressions using recognition-based analysis of image sequences. in Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). vol. 1352, Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 1352, Springer Verlag, pp. 703-710, 3rd Asian Conference on Computer Vision, ACCV 1998, Hong Kong, Hong Kong, 98/1/8.
Otsuka T, Ohya J. Converting facial expressions using recognition-based analysis of image sequences. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). Vol. 1352. Springer Verlag. 1997. p. 703-710. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).
Otsuka, Takahiro ; Ohya, Jun. / Converting facial expressions using recognition-based analysis of image sequences. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). Vol. 1352 Springer Verlag, 1997. pp. 703-710 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).
@inproceedings{bb6fd13098dc494896dd8031d1aa4d52,
title = "Converting facial expressions using recognition-based analysis of image sequences",
abstract = "A method for converting one person’s facial expression into another person’s is proposed. The sequence of the feature vector for each expression is modeled by using HMM with the hidden states corresponding to the different muscle conditions (relaxed, contracting, and the end of contraction). The probability of each state is evaluated for each frame and the contraction rate of each muscle is obtained from the probability of each state using a matrix representing the characteristics of other people’s expressions. The experiments showed the superior realism of the expression generated by our proposed method.",
author = "Takahiro Otsuka and Jun Ohya",
year = "1997",
language = "English",
isbn = "3540639314",
volume = "1352",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
publisher = "Springer Verlag",
pages = "703--710",
booktitle = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

}

TY - GEN

T1 - Converting facial expressions using recognition-based analysis of image sequences

AU - Otsuka, Takahiro

AU - Ohya, Jun

PY - 1997

Y1 - 1997

N2 - A method for converting one person’s facial expression into another person’s is proposed. The sequence of the feature vector for each expression is modeled by using HMM with the hidden states corresponding to the different muscle conditions (relaxed, contracting, and the end of contraction). The probability of each state is evaluated for each frame and the contraction rate of each muscle is obtained from the probability of each state using a matrix representing the characteristics of other people’s expressions. The experiments showed the superior realism of the expression generated by our proposed method.

AB - A method for converting one person’s facial expression into another person’s is proposed. The sequence of the feature vector for each expression is modeled by using HMM with the hidden states corresponding to the different muscle conditions (relaxed, contracting, and the end of contraction). The probability of each state is evaluated for each frame and the contraction rate of each muscle is obtained from the probability of each state using a matrix representing the characteristics of other people’s expressions. The experiments showed the superior realism of the expression generated by our proposed method.

UR - http://www.scopus.com/inward/record.url?scp=84947590447&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84947590447&partnerID=8YFLogxK

M3 - Conference contribution

SN - 3540639314

SN - 9783540639312

VL - 1352

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 703

EP - 710

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

PB - Springer Verlag

ER -