Multi-band Modelling of Appearance

Mikkel Bille Stegmann; Rasmus Larsen

doi:10.1016/S0262-8856(02)00126-9

Multi-band Modelling of Appearance

Mikkel Bille Stegmann, Rasmus Larsen

Research output: Contribution to journal › Conference article › Research › peer-review

Abstract

Earlier work has demonstrated generative models capable of synthesising near photo-realistic grey-scale images of objects. These models have been augmented with colour information, and recently with edge information. This paper extends the active appearance model framework by modelling the appearance of both derived feature bands and an intensity band. As a special case of feature-band augmented appearance modelling we propose a dedicated representation with applications to face segmentation. The representation addresses a major problem within face recognition by lowering the sensitivity to lighting conditions. Results show that the localisation accuracy of facial features is considerably increased using this appearance representation under diffuse and directional lighting and at multiple scales.

Original language	English
Journal	Image and Vision Computing
Volume	21
Issue number	1
Pages (from-to)	61-67
ISSN	0262-8856
DOIs	https://doi.org/10.1016/S0262-8856(02)00126-9
Publication status	Published - 2003
Event	1st International Workshop on Generative-Model-Based Vision - Copenhagen, Denmark Duration: 2 Jun 2002 → … Conference number: 1

Conference

Conference	1st International Workshop on Generative-Model-Based Vision
Number	1
Country/Territory	Denmark
City	Copenhagen
Period	02/06/2002 → …

Keywords

Generative model
Active appearance model
Lighting invariance
Face recognition
Segmentation

Access to Document

10.1016/S0262-8856(02)00126-9

http://www.imm.dtu.dk/pubdb/p.php?1421

OpenUrl availability

Full text

Cite this

@inproceedings{1ec0a29641f747d9b19b59c9503d841e,

title = "Multi-band Modelling of Appearance",

abstract = "Earlier work has demonstrated generative models capable of synthesising near photo-realistic grey-scale images of objects. These models have been augmented with colour information, and recently with edge information. This paper extends the active appearance model framework by modelling the appearance of both derived feature bands and an intensity band. As a special case of feature-band augmented appearance modelling we propose a dedicated representation with applications to face segmentation. The representation addresses a major problem within face recognition by lowering the sensitivity to lighting conditions. Results show that the localisation accuracy of facial features is considerably increased using this appearance representation under diffuse and directional lighting and at multiple scales.",

keywords = "Generative model; Active appearance model; Lighting invariance; Face recognition; Segmentation, Generative model, Active appearance model, Lighting invariance, Face recognition, Segmentation",

author = "Stegmann, {Mikkel Bille} and Rasmus Larsen",

year = "2003",

doi = "10.1016/S0262-8856(02)00126-9",

language = "English",

volume = "21",

pages = "61--67",

journal = "Image and Vision Computing",

issn = "0262-8856",

publisher = "Elsevier",

number = "1",

note = "1st International Workshop on Generative-Model-Based Vision, GMBV 2002 ; Conference date: 02-06-2002",

}

TY - GEN

T1 - Multi-band Modelling of Appearance

AU - Stegmann, Mikkel Bille

AU - Larsen, Rasmus

N1 - Conference code: 1

PY - 2003

Y1 - 2003

N2 - Earlier work has demonstrated generative models capable of synthesising near photo-realistic grey-scale images of objects. These models have been augmented with colour information, and recently with edge information. This paper extends the active appearance model framework by modelling the appearance of both derived feature bands and an intensity band. As a special case of feature-band augmented appearance modelling we propose a dedicated representation with applications to face segmentation. The representation addresses a major problem within face recognition by lowering the sensitivity to lighting conditions. Results show that the localisation accuracy of facial features is considerably increased using this appearance representation under diffuse and directional lighting and at multiple scales.

AB - Earlier work has demonstrated generative models capable of synthesising near photo-realistic grey-scale images of objects. These models have been augmented with colour information, and recently with edge information. This paper extends the active appearance model framework by modelling the appearance of both derived feature bands and an intensity band. As a special case of feature-band augmented appearance modelling we propose a dedicated representation with applications to face segmentation. The representation addresses a major problem within face recognition by lowering the sensitivity to lighting conditions. Results show that the localisation accuracy of facial features is considerably increased using this appearance representation under diffuse and directional lighting and at multiple scales.

KW - Generative model; Active appearance model; Lighting invariance; Face recognition; Segmentation

KW - Generative model

KW - Active appearance model

KW - Lighting invariance

KW - Face recognition

KW - Segmentation

U2 - 10.1016/S0262-8856(02)00126-9

DO - 10.1016/S0262-8856(02)00126-9

M3 - Conference article

SN - 0262-8856

VL - 21

SP - 61

EP - 67

JO - Image and Vision Computing

JF - Image and Vision Computing

IS - 1

T2 - 1st International Workshop on Generative-Model-Based Vision

Y2 - 2 June 2002

ER -

Multi-band Modelling of Appearance

Abstract

Conference

Keywords

Access to Document

OpenUrl availability

Fingerprint

Cite this