A Joint Discriminative Generative Model for Deformable Model Construction and Classification

Ioannis Marras, Symeon Nikitidis, Stefanos Zafeiriou, Maja Pantic

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

Abstract

Discriminative classification models have been successfully applied for various computer vision tasks such as object and face detection and recognition. However, deformations can change objects coordinate space and perturb robust similarity measurement, which is the essence of all classification algorithms. The common approach to deal with deformations is either to seek for deformation invariant features or to develop models that describe objects deformations. However, the former approach requires a huge amount of data and a good amount of engineering to be properly trained, while the latter require considerable human effort in the form of carefully annotated data. In this paper, we propose a method that jointly learns with minimal human intervention a generative deformable model using only a simple shape model of the object and images automatically downloaded from the Internet, and also extracts features appropriate for classification. The proposed algorithm is applied on various classification problems such as 'in-thewild' face recognition, gender classification and eye glasses detection on data retrieved by querying into a web image search engine. We demonstrate that not only it outperforms other automatic methods by large margins, but also performs comparably with supervised methods trained on thousands of manually annotated data.

Original languageEnglish
Title of host publication12th IEEE International Conference on Automatic Face & Gesture Recognition (FG 2017)
PublisherIEEE
Pages127-134
Number of pages8
ISBN (Electronic)9781509040230
ISBN (Print)978-1-5090-4024-7
DOIs
Publication statusPublished - 28 Jun 2017
Event12th IEEE International Conference on Automatic Face and Gesture Recognition, FG 2017 - Washington, United States
Duration: 30 May 20173 Jun 2017
Conference number: 12
http://www.fg2017.org/

Conference

Conference12th IEEE International Conference on Automatic Face and Gesture Recognition, FG 2017
Abbreviated titleFG
CountryUnited States
CityWashington
Period30/05/173/06/17
Internet address

Fingerprint

Face recognition
Search engines
World Wide Web
Computer vision
Internet
Glass

Cite this

Marras, I., Nikitidis, S., Zafeiriou, S., & Pantic, M. (2017). A Joint Discriminative Generative Model for Deformable Model Construction and Classification. In 12th IEEE International Conference on Automatic Face & Gesture Recognition (FG 2017) (pp. 127-134). [7961732] IEEE. https://doi.org/10.1109/FG.2017.24
Marras, Ioannis ; Nikitidis, Symeon ; Zafeiriou, Stefanos ; Pantic, Maja. / A Joint Discriminative Generative Model for Deformable Model Construction and Classification. 12th IEEE International Conference on Automatic Face & Gesture Recognition (FG 2017). IEEE, 2017. pp. 127-134
@inproceedings{f0d89c35ec6343b093d7b95d8ed398e3,
title = "A Joint Discriminative Generative Model for Deformable Model Construction and Classification",
abstract = "Discriminative classification models have been successfully applied for various computer vision tasks such as object and face detection and recognition. However, deformations can change objects coordinate space and perturb robust similarity measurement, which is the essence of all classification algorithms. The common approach to deal with deformations is either to seek for deformation invariant features or to develop models that describe objects deformations. However, the former approach requires a huge amount of data and a good amount of engineering to be properly trained, while the latter require considerable human effort in the form of carefully annotated data. In this paper, we propose a method that jointly learns with minimal human intervention a generative deformable model using only a simple shape model of the object and images automatically downloaded from the Internet, and also extracts features appropriate for classification. The proposed algorithm is applied on various classification problems such as 'in-thewild' face recognition, gender classification and eye glasses detection on data retrieved by querying into a web image search engine. We demonstrate that not only it outperforms other automatic methods by large margins, but also performs comparably with supervised methods trained on thousands of manually annotated data.",
author = "Ioannis Marras and Symeon Nikitidis and Stefanos Zafeiriou and Maja Pantic",
year = "2017",
month = "6",
day = "28",
doi = "10.1109/FG.2017.24",
language = "English",
isbn = "978-1-5090-4024-7",
pages = "127--134",
booktitle = "12th IEEE International Conference on Automatic Face & Gesture Recognition (FG 2017)",
publisher = "IEEE",
address = "United States",

}

Marras, I, Nikitidis, S, Zafeiriou, S & Pantic, M 2017, A Joint Discriminative Generative Model for Deformable Model Construction and Classification. in 12th IEEE International Conference on Automatic Face & Gesture Recognition (FG 2017)., 7961732, IEEE, pp. 127-134, 12th IEEE International Conference on Automatic Face and Gesture Recognition, FG 2017, Washington, United States, 30/05/17. https://doi.org/10.1109/FG.2017.24

A Joint Discriminative Generative Model for Deformable Model Construction and Classification. / Marras, Ioannis; Nikitidis, Symeon; Zafeiriou, Stefanos; Pantic, Maja.

12th IEEE International Conference on Automatic Face & Gesture Recognition (FG 2017). IEEE, 2017. p. 127-134 7961732.

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

TY - GEN

T1 - A Joint Discriminative Generative Model for Deformable Model Construction and Classification

AU - Marras, Ioannis

AU - Nikitidis, Symeon

AU - Zafeiriou, Stefanos

AU - Pantic, Maja

PY - 2017/6/28

Y1 - 2017/6/28

N2 - Discriminative classification models have been successfully applied for various computer vision tasks such as object and face detection and recognition. However, deformations can change objects coordinate space and perturb robust similarity measurement, which is the essence of all classification algorithms. The common approach to deal with deformations is either to seek for deformation invariant features or to develop models that describe objects deformations. However, the former approach requires a huge amount of data and a good amount of engineering to be properly trained, while the latter require considerable human effort in the form of carefully annotated data. In this paper, we propose a method that jointly learns with minimal human intervention a generative deformable model using only a simple shape model of the object and images automatically downloaded from the Internet, and also extracts features appropriate for classification. The proposed algorithm is applied on various classification problems such as 'in-thewild' face recognition, gender classification and eye glasses detection on data retrieved by querying into a web image search engine. We demonstrate that not only it outperforms other automatic methods by large margins, but also performs comparably with supervised methods trained on thousands of manually annotated data.

AB - Discriminative classification models have been successfully applied for various computer vision tasks such as object and face detection and recognition. However, deformations can change objects coordinate space and perturb robust similarity measurement, which is the essence of all classification algorithms. The common approach to deal with deformations is either to seek for deformation invariant features or to develop models that describe objects deformations. However, the former approach requires a huge amount of data and a good amount of engineering to be properly trained, while the latter require considerable human effort in the form of carefully annotated data. In this paper, we propose a method that jointly learns with minimal human intervention a generative deformable model using only a simple shape model of the object and images automatically downloaded from the Internet, and also extracts features appropriate for classification. The proposed algorithm is applied on various classification problems such as 'in-thewild' face recognition, gender classification and eye glasses detection on data retrieved by querying into a web image search engine. We demonstrate that not only it outperforms other automatic methods by large margins, but also performs comparably with supervised methods trained on thousands of manually annotated data.

UR - http://www.scopus.com/inward/record.url?scp=85026303597&partnerID=8YFLogxK

U2 - 10.1109/FG.2017.24

DO - 10.1109/FG.2017.24

M3 - Conference contribution

SN - 978-1-5090-4024-7

SP - 127

EP - 134

BT - 12th IEEE International Conference on Automatic Face & Gesture Recognition (FG 2017)

PB - IEEE

ER -

Marras I, Nikitidis S, Zafeiriou S, Pantic M. A Joint Discriminative Generative Model for Deformable Model Construction and Classification. In 12th IEEE International Conference on Automatic Face & Gesture Recognition (FG 2017). IEEE. 2017. p. 127-134. 7961732 https://doi.org/10.1109/FG.2017.24