Multi-stream Gaussian mixture model based facial feature localization

Kenichi Kumatani; Hazim K. Ekenel; Hua Gao; Rainer Stiefelhagen; Aytül Erçil

doi:10.1109/SIU.2008.4632752

Multi-stream Gaussian mixture model based facial feature localization

Kenichi Kumatani, Hazim K. Ekenel, Hua Gao, Rainer Stiefelhagen, Aytül Erçil

Computer Engineering

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

Abstract

This paper presents a new facial feature localization system which estimates positions of eyes, nose and mouth corners simultaneously. In contrast to conventional systems, we use the multi-stream Gaussian mixture model (GMM) framework in order to represent structural and appearance information of facial features. We construct a GMM for the region of each facial feature, where the principal component analysis is used to extract each facial feature. We also build a GMM which represents the structural information of a face, relative positions of facial features. Those models are combined based on the multi-stream framework. It can reduce the computation time to search region of interest (ROI). We demonstrate the effectiveness of our algorithm through experiments on the BioID Face Database.

Original language	English (US)
Title of host publication	2008 IEEE 16th Signal Processing, Communication and Applications Conference, SIU
DOIs	https://doi.org/10.1109/SIU.2008.4632752
State	Published - 2008
Event	2008 IEEE 16th Signal Processing, Communication and Applications Conference, SIU - Aydin, Turkey Duration: Apr 20 2008 → Apr 22 2008

Publication series

Name	2008 IEEE 16th Signal Processing, Communication and Applications Conference, SIU

Conference

Conference	2008 IEEE 16th Signal Processing, Communication and Applications Conference, SIU
Country/Territory	Turkey
City	Aydin
Period	4/20/08 → 4/22/08

ASJC Scopus subject areas

Signal Processing
Electrical and Electronic Engineering
Communication

Access to Document

10.1109/SIU.2008.4632752

Cite this

Kumatani, K., Ekenel, H. K., Gao, H., Stiefelhagen, R., & Erçil, A. (2008). Multi-stream Gaussian mixture model based facial feature localization. In 2008 IEEE 16th Signal Processing, Communication and Applications Conference, SIU Article 4632752 (2008 IEEE 16th Signal Processing, Communication and Applications Conference, SIU). https://doi.org/10.1109/SIU.2008.4632752

Multi-stream Gaussian mixture model based facial feature localization. / Kumatani, Kenichi; Ekenel, Hazim K.; Gao, Hua et al.
2008 IEEE 16th Signal Processing, Communication and Applications Conference, SIU. 2008. 4632752 (2008 IEEE 16th Signal Processing, Communication and Applications Conference, SIU).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

Kumatani, K, Ekenel, HK, Gao, H, Stiefelhagen, R & Erçil, A 2008, Multi-stream Gaussian mixture model based facial feature localization. in 2008 IEEE 16th Signal Processing, Communication and Applications Conference, SIU., 4632752, 2008 IEEE 16th Signal Processing, Communication and Applications Conference, SIU, 2008 IEEE 16th Signal Processing, Communication and Applications Conference, SIU, Aydin, Turkey, 4/20/08. https://doi.org/10.1109/SIU.2008.4632752

@inproceedings{664593ac7ce842fda219a558a60b7ecf,

title = "Multi-stream Gaussian mixture model based facial feature localization",

abstract = "This paper presents a new facial feature localization system which estimates positions of eyes, nose and mouth corners simultaneously. In contrast to conventional systems, we use the multi-stream Gaussian mixture model (GMM) framework in order to represent structural and appearance information of facial features. We construct a GMM for the region of each facial feature, where the principal component analysis is used to extract each facial feature. We also build a GMM which represents the structural information of a face, relative positions of facial features. Those models are combined based on the multi-stream framework. It can reduce the computation time to search region of interest (ROI). We demonstrate the effectiveness of our algorithm through experiments on the BioID Face Database.",

author = "Kenichi Kumatani and Ekenel, {Hazim K.} and Hua Gao and Rainer Stiefelhagen and Ayt{\"u}l Er{\c c}il",

year = "2008",

doi = "10.1109/SIU.2008.4632752",

language = "English (US)",

isbn = "9781424419999",

series = "2008 IEEE 16th Signal Processing, Communication and Applications Conference, SIU",

booktitle = "2008 IEEE 16th Signal Processing, Communication and Applications Conference, SIU",

note = "2008 IEEE 16th Signal Processing, Communication and Applications Conference, SIU ; Conference date: 20-04-2008 Through 22-04-2008",

}

TY - GEN

T1 - Multi-stream Gaussian mixture model based facial feature localization

AU - Kumatani, Kenichi

AU - Ekenel, Hazim K.

AU - Gao, Hua

AU - Stiefelhagen, Rainer

AU - Erçil, Aytül

PY - 2008

Y1 - 2008

N2 - This paper presents a new facial feature localization system which estimates positions of eyes, nose and mouth corners simultaneously. In contrast to conventional systems, we use the multi-stream Gaussian mixture model (GMM) framework in order to represent structural and appearance information of facial features. We construct a GMM for the region of each facial feature, where the principal component analysis is used to extract each facial feature. We also build a GMM which represents the structural information of a face, relative positions of facial features. Those models are combined based on the multi-stream framework. It can reduce the computation time to search region of interest (ROI). We demonstrate the effectiveness of our algorithm through experiments on the BioID Face Database.

AB - This paper presents a new facial feature localization system which estimates positions of eyes, nose and mouth corners simultaneously. In contrast to conventional systems, we use the multi-stream Gaussian mixture model (GMM) framework in order to represent structural and appearance information of facial features. We construct a GMM for the region of each facial feature, where the principal component analysis is used to extract each facial feature. We also build a GMM which represents the structural information of a face, relative positions of facial features. Those models are combined based on the multi-stream framework. It can reduce the computation time to search region of interest (ROI). We demonstrate the effectiveness of our algorithm through experiments on the BioID Face Database.

UR - http://www.scopus.com/inward/record.url?scp=56449090608&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=56449090608&partnerID=8YFLogxK

U2 - 10.1109/SIU.2008.4632752

DO - 10.1109/SIU.2008.4632752

M3 - Conference contribution

AN - SCOPUS:56449090608

SN - 9781424419999

T3 - 2008 IEEE 16th Signal Processing, Communication and Applications Conference, SIU

BT - 2008 IEEE 16th Signal Processing, Communication and Applications Conference, SIU

T2 - 2008 IEEE 16th Signal Processing, Communication and Applications Conference, SIU

Y2 - 20 April 2008 through 22 April 2008

ER -

Multi-stream Gaussian mixture model based facial feature localization

Abstract

Publication series

Conference

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this