QCompere @ REPERE 2013

Hervé Bredin; Johann Poignant; Guillaume Fortier; Makarand Tapaswi; Viet Bac Le; Anindya Roy; Claude Barras; Sophie Rosset; Achintya Sarkar; Qian Yang; Hua Gao; Alexis Mignon; Jakob Verbeek; Laurent Besacier; Georges Quénot; Hazim Kemal Ekenel; Rainer Stiefelhagen

QCompere @ REPERE 2013

Hervé Bredin, Johann Poignant, Guillaume Fortier, Makarand Tapaswi, Viet Bac Le, Anindya Roy, Claude Barras, Sophie Rosset, Achintya Sarkar, Qian Yang, Hua Gao, Alexis Mignon, Jakob Verbeek, Laurent Besacier, Georges Quénot, Hazim Kemal Ekenel, Rainer Stiefelhagen

Computer Engineering

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

Abstract

We describe QCompere consortium submissions to the REPERE 2013 evaluation campaign. The REPERE challenge aims at gathering four communities (face recognition, speaker identification, optical character recognition and named entity detection) towards the same goal: multimodal person recognition in TV broadcast. First, four mono-modal components are introduced (one for each foregoing community) constituting the elementary building blocks of our various submissions. Then, depending on the target modality (speaker or face recognition) and on the task (supervised or unsupervised recognition), four different fusion techniques are introduced: they can be summarized as propagation-, classifier-, rule- or graph-based approaches. Finally, their performance is evaluated on REPERE 2013 test set and their advantages and limitations are discussed.

Original language	English (US)
Pages (from-to)	49-54
Number of pages	6
Journal	CEUR Workshop Proceedings
Volume	1012
State	Published - 2013
Event	1st Workshop on Speech, Language and Audio in Multimedia, SLAM 2013 - Marseille, France Duration: Aug 22 2013 → Aug 23 2013

Keywords

Face recognition
Multimodal fusion
Named entity detection
Speaker identification
Video optical character recognition

ASJC Scopus subject areas

General Computer Science

Cite this

@article{04bf1f04fbc4473f88dcf5e51a62dac0,

title = "QCompere @ REPERE 2013",

abstract = "We describe QCompere consortium submissions to the REPERE 2013 evaluation campaign. The REPERE challenge aims at gathering four communities (face recognition, speaker identification, optical character recognition and named entity detection) towards the same goal: multimodal person recognition in TV broadcast. First, four mono-modal components are introduced (one for each foregoing community) constituting the elementary building blocks of our various submissions. Then, depending on the target modality (speaker or face recognition) and on the task (supervised or unsupervised recognition), four different fusion techniques are introduced: they can be summarized as propagation-, classifier-, rule- or graph-based approaches. Finally, their performance is evaluated on REPERE 2013 test set and their advantages and limitations are discussed.",

keywords = "Face recognition, Multimodal fusion, Named entity detection, Speaker identification, Video optical character recognition",

author = "Herv{\'e} Bredin and Johann Poignant and Guillaume Fortier and Makarand Tapaswi and Le, {Viet Bac} and Anindya Roy and Claude Barras and Sophie Rosset and Achintya Sarkar and Qian Yang and Hua Gao and Alexis Mignon and Jakob Verbeek and Laurent Besacier and Georges Qu{\'e}not and Ekenel, {Hazim Kemal} and Rainer Stiefelhagen",

note = "Publisher Copyright: Copyright {\textcopyright} 2013 for the individual papers by the papers' authors.; 1st Workshop on Speech, Language and Audio in Multimedia, SLAM 2013 ; Conference date: 22-08-2013 Through 23-08-2013",

year = "2013",

language = "English (US)",

volume = "1012",

pages = "49--54",

journal = "CEUR Workshop Proceedings",

issn = "1613-0073",

publisher = "CEUR-WS",

}

TY - JOUR

T1 - QCompere @ REPERE 2013

AU - Bredin, Hervé

AU - Poignant, Johann

AU - Fortier, Guillaume

AU - Tapaswi, Makarand

AU - Le, Viet Bac

AU - Roy, Anindya

AU - Barras, Claude

AU - Rosset, Sophie

AU - Sarkar, Achintya

AU - Yang, Qian

AU - Gao, Hua

AU - Mignon, Alexis

AU - Verbeek, Jakob

AU - Besacier, Laurent

AU - Quénot, Georges

AU - Ekenel, Hazim Kemal

AU - Stiefelhagen, Rainer

PY - 2013

Y1 - 2013

N2 - We describe QCompere consortium submissions to the REPERE 2013 evaluation campaign. The REPERE challenge aims at gathering four communities (face recognition, speaker identification, optical character recognition and named entity detection) towards the same goal: multimodal person recognition in TV broadcast. First, four mono-modal components are introduced (one for each foregoing community) constituting the elementary building blocks of our various submissions. Then, depending on the target modality (speaker or face recognition) and on the task (supervised or unsupervised recognition), four different fusion techniques are introduced: they can be summarized as propagation-, classifier-, rule- or graph-based approaches. Finally, their performance is evaluated on REPERE 2013 test set and their advantages and limitations are discussed.

AB - We describe QCompere consortium submissions to the REPERE 2013 evaluation campaign. The REPERE challenge aims at gathering four communities (face recognition, speaker identification, optical character recognition and named entity detection) towards the same goal: multimodal person recognition in TV broadcast. First, four mono-modal components are introduced (one for each foregoing community) constituting the elementary building blocks of our various submissions. Then, depending on the target modality (speaker or face recognition) and on the task (supervised or unsupervised recognition), four different fusion techniques are introduced: they can be summarized as propagation-, classifier-, rule- or graph-based approaches. Finally, their performance is evaluated on REPERE 2013 test set and their advantages and limitations are discussed.

KW - Face recognition

KW - Multimodal fusion

KW - Named entity detection

KW - Speaker identification

KW - Video optical character recognition

UR - http://www.scopus.com/inward/record.url?scp=84903747023&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84903747023&partnerID=8YFLogxK

M3 - Conference article

AN - SCOPUS:84903747023

SN - 1613-0073

VL - 1012

SP - 49

EP - 54

JO - CEUR Workshop Proceedings

JF - CEUR Workshop Proceedings

T2 - 1st Workshop on Speech, Language and Audio in Multimedia, SLAM 2013

Y2 - 22 August 2013 through 23 August 2013

ER -

QCompere @ REPERE 2013

Abstract

Keywords

ASJC Scopus subject areas

Other files and links

Fingerprint

Cite this