Shared Gaussian Process Latent Variable Model for Multi-view Facial Expression Recognition

Stefanos Eleftheriadis, Ognjen Rudovic, Maja Pantic

Research output: Chapter in Book/Report/Conference proceedingConference contributionAcademicpeer-review

13 Citations (Scopus)
15 Downloads (Pure)

Abstract

Facial-expression data often appear in multiple views either due to head-movements or the camera position. Existing methods for multi-view facial expression recognition perform classification of the target expressions either by using classifiers learned separately for each view or by using a single classifier learned for all views. However, these approaches do not explore the fact that multi-view facial expression data are different manifestations of the same facial-expression-related latent content. To this end, we propose a Shared Gaussian Process Latent Variable Model (SGPLVM) for classification of multi-view facial expression data. In this model, we first learn a discriminative manifold shared by multiple views of facial expressions, and then apply a (single) facial expression classifier, based on k-Nearest-Neighbours (kNN), to the shared manifold. In our experiments on the MultiPIE database, containing real images of facial expressions in multiple views, we show that the proposed model outperforms the state-of-the-art models for multi-view facial expression recognition.
Original languageEnglish
Title of host publicationAdvances in Visual Computing
Subtitle of host publication9th International Symposium, ISVC 2013, Rethymnon, Crete, Greece, July 29-31, 2013. Proceedings
EditorsGeorge Bebis, Richard Boyle, Bahram Parvin, Darko Koracin
Place of PublicationBerlin Heidelberg
PublisherSpringer
Pages527-538
Number of pages12
VolumePart I
ISBN (Print)978-3-642-41913-3
DOIs
Publication statusPublished - Jul 2013
Event9th International Symposium on Visual Computing, ISVC 2013 - Rethymnon, Crete, Greece
Duration: 29 Jul 201331 Jul 2013
Conference number: 9

Publication series

NameLecture Notes in Computer Science
PublisherSpringer
Volume8033
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

Conference9th International Symposium on Visual Computing, ISVC 2013
Abbreviated titleISVC
CountryGreece
CityRethymnon, Crete
Period29/07/1331/07/13

Fingerprint

Classifiers
Cameras
Experiments

Keywords

  • EWI-24327
  • METIS-302654
  • IR-89368
  • HMI-HF: Human Factors

Cite this

Eleftheriadis, S., Rudovic, O., & Pantic, M. (2013). Shared Gaussian Process Latent Variable Model for Multi-view Facial Expression Recognition. In G. Bebis, R. Boyle, B. Parvin, & D. Koracin (Eds.), Advances in Visual Computing: 9th International Symposium, ISVC 2013, Rethymnon, Crete, Greece, July 29-31, 2013. Proceedings (Vol. Part I, pp. 527-538). (Lecture Notes in Computer Science; Vol. 8033). Berlin Heidelberg: Springer. https://doi.org/10.1007/978-3-642-41914-0_52
Eleftheriadis, Stefanos ; Rudovic, Ognjen ; Pantic, Maja . / Shared Gaussian Process Latent Variable Model for Multi-view Facial Expression Recognition. Advances in Visual Computing: 9th International Symposium, ISVC 2013, Rethymnon, Crete, Greece, July 29-31, 2013. Proceedings. editor / George Bebis ; Richard Boyle ; Bahram Parvin ; Darko Koracin. Vol. Part I Berlin Heidelberg : Springer, 2013. pp. 527-538 (Lecture Notes in Computer Science).
@inproceedings{c8e0ae33c68b4adf97368ef75a84b48c,
title = "Shared Gaussian Process Latent Variable Model for Multi-view Facial Expression Recognition",
abstract = "Facial-expression data often appear in multiple views either due to head-movements or the camera position. Existing methods for multi-view facial expression recognition perform classification of the target expressions either by using classifiers learned separately for each view or by using a single classifier learned for all views. However, these approaches do not explore the fact that multi-view facial expression data are different manifestations of the same facial-expression-related latent content. To this end, we propose a Shared Gaussian Process Latent Variable Model (SGPLVM) for classification of multi-view facial expression data. In this model, we first learn a discriminative manifold shared by multiple views of facial expressions, and then apply a (single) facial expression classifier, based on k-Nearest-Neighbours (kNN), to the shared manifold. In our experiments on the MultiPIE database, containing real images of facial expressions in multiple views, we show that the proposed model outperforms the state-of-the-art models for multi-view facial expression recognition.",
keywords = "EWI-24327, METIS-302654, IR-89368, HMI-HF: Human Factors",
author = "Stefanos Eleftheriadis and Ognjen Rudovic and Maja Pantic",
year = "2013",
month = "7",
doi = "10.1007/978-3-642-41914-0_52",
language = "English",
isbn = "978-3-642-41913-3",
volume = "Part I",
series = "Lecture Notes in Computer Science",
publisher = "Springer",
pages = "527--538",
editor = "George Bebis and Richard Boyle and Bahram Parvin and Darko Koracin",
booktitle = "Advances in Visual Computing",

}

Eleftheriadis, S, Rudovic, O & Pantic, M 2013, Shared Gaussian Process Latent Variable Model for Multi-view Facial Expression Recognition. in G Bebis, R Boyle, B Parvin & D Koracin (eds), Advances in Visual Computing: 9th International Symposium, ISVC 2013, Rethymnon, Crete, Greece, July 29-31, 2013. Proceedings. vol. Part I, Lecture Notes in Computer Science, vol. 8033, Springer, Berlin Heidelberg, pp. 527-538, 9th International Symposium on Visual Computing, ISVC 2013, Rethymnon, Crete, Greece, 29/07/13. https://doi.org/10.1007/978-3-642-41914-0_52

Shared Gaussian Process Latent Variable Model for Multi-view Facial Expression Recognition. / Eleftheriadis, Stefanos; Rudovic, Ognjen; Pantic, Maja .

Advances in Visual Computing: 9th International Symposium, ISVC 2013, Rethymnon, Crete, Greece, July 29-31, 2013. Proceedings. ed. / George Bebis; Richard Boyle; Bahram Parvin; Darko Koracin. Vol. Part I Berlin Heidelberg : Springer, 2013. p. 527-538 (Lecture Notes in Computer Science; Vol. 8033).

Research output: Chapter in Book/Report/Conference proceedingConference contributionAcademicpeer-review

TY - GEN

T1 - Shared Gaussian Process Latent Variable Model for Multi-view Facial Expression Recognition

AU - Eleftheriadis, Stefanos

AU - Rudovic, Ognjen

AU - Pantic, Maja

PY - 2013/7

Y1 - 2013/7

N2 - Facial-expression data often appear in multiple views either due to head-movements or the camera position. Existing methods for multi-view facial expression recognition perform classification of the target expressions either by using classifiers learned separately for each view or by using a single classifier learned for all views. However, these approaches do not explore the fact that multi-view facial expression data are different manifestations of the same facial-expression-related latent content. To this end, we propose a Shared Gaussian Process Latent Variable Model (SGPLVM) for classification of multi-view facial expression data. In this model, we first learn a discriminative manifold shared by multiple views of facial expressions, and then apply a (single) facial expression classifier, based on k-Nearest-Neighbours (kNN), to the shared manifold. In our experiments on the MultiPIE database, containing real images of facial expressions in multiple views, we show that the proposed model outperforms the state-of-the-art models for multi-view facial expression recognition.

AB - Facial-expression data often appear in multiple views either due to head-movements or the camera position. Existing methods for multi-view facial expression recognition perform classification of the target expressions either by using classifiers learned separately for each view or by using a single classifier learned for all views. However, these approaches do not explore the fact that multi-view facial expression data are different manifestations of the same facial-expression-related latent content. To this end, we propose a Shared Gaussian Process Latent Variable Model (SGPLVM) for classification of multi-view facial expression data. In this model, we first learn a discriminative manifold shared by multiple views of facial expressions, and then apply a (single) facial expression classifier, based on k-Nearest-Neighbours (kNN), to the shared manifold. In our experiments on the MultiPIE database, containing real images of facial expressions in multiple views, we show that the proposed model outperforms the state-of-the-art models for multi-view facial expression recognition.

KW - EWI-24327

KW - METIS-302654

KW - IR-89368

KW - HMI-HF: Human Factors

U2 - 10.1007/978-3-642-41914-0_52

DO - 10.1007/978-3-642-41914-0_52

M3 - Conference contribution

SN - 978-3-642-41913-3

VL - Part I

T3 - Lecture Notes in Computer Science

SP - 527

EP - 538

BT - Advances in Visual Computing

A2 - Bebis, George

A2 - Boyle, Richard

A2 - Parvin, Bahram

A2 - Koracin, Darko

PB - Springer

CY - Berlin Heidelberg

ER -

Eleftheriadis S, Rudovic O, Pantic M. Shared Gaussian Process Latent Variable Model for Multi-view Facial Expression Recognition. In Bebis G, Boyle R, Parvin B, Koracin D, editors, Advances in Visual Computing: 9th International Symposium, ISVC 2013, Rethymnon, Crete, Greece, July 29-31, 2013. Proceedings. Vol. Part I. Berlin Heidelberg: Springer. 2013. p. 527-538. (Lecture Notes in Computer Science). https://doi.org/10.1007/978-3-642-41914-0_52