Fusion of Speech, Faces and Text for Person Identification in TV Broadcast

Hervé Bredin; Johann Poignant; Makarand Tapaswi; Guillaume Fortier; Viet Bac Le; Thibault Napoléon; Gao Hua; Claude Barras; Sophie Rosset; Laurent Besacier; Jakob Verbeek; Georges Quénot; Frédéric Jurie; Ekenel Hazim Kemal

Fusion of Speech, Faces and Text for Person Identification in TV Broadcast

Hervé Bredin ¹ Johann Poignant ² Makarand Tapaswi ³ Guillaume Fortier ⁴ Viet Bac Le ⁵ Thibault Napoléon ⁶ Gao Hua ³ Claude Barras ¹ Sophie Rosset ¹ Laurent Besacier ⁷ Jakob Verbeek ⁴ Georges Quénot ^{8, *} Frédéric Jurie ^{4, 9} Ekenel Hazim Kemal ³

* Corresponding author

1 LIMSI - Laboratoire d'Informatique pour la Mécanique et les Sciences de l'Ingénieur

2 LIG - Laboratoire d'Informatique de Grenoble

3 KIT - Karlsruhe Institute of Technology

4 LEAR - Learning and recognition in vision

Inria Grenoble - Rhône-Alpes, LJK - Laboratoire Jean Kuntzmann, INPG - Institut National Polytechnique de Grenoble

5 Vocapia Research [Orsay]

6 GREYC - Groupe de Recherche en Informatique, Image, Automatique et Instrumentation de Caen

7 GETALP - Groupe d’Étude en Traduction Automatique/Traitement Automatisé des Langues et de la Parole

LIG - Laboratoire d'Informatique de Grenoble

8 MRIM - Modélisation et Recherche d’Information Multimédia [Grenoble]

Inria - Institut National de Recherche en Informatique et en Automatique, LIG - Laboratoire d'Informatique de Grenoble

9 Equipe Image - Laboratoire GREYC - UMR6072

GREYC - Groupe de Recherche en Informatique, Image, Automatique et Instrumentation de Caen

Abstract : The REPERE challenge is a project aiming at the evaluation of systems for supervised and unsupervised multimodal recognition of people in TV broadcast. In this paper, we describe, evaluate and discuss QCompere consortium submissions to the 2012 REPERE evaluation campaign dry-run. Speaker identification (and face recognition) can be greatly improved when combined with name detection through video optical character recognition. Moreover, we show that unsupervised multimodal person recognition systems can achieve performance nearly as good as supervised monomodal ones (with several hundreds of identity models).

Document type :

Conference papers

Domain :

Computer Science [cs] / Machine Learning [cs.LG]

Computer Science [cs] / Information Retrieval [cs.IR]

Complete list of metadatas

Cited literature [12 references] Display Hide Download

https://hal.inria.fr/hal-00722884
Contributor : Thoth Team <>
Submitted on : Monday, August 6, 2012 - 10:34:55 AM
Last modification on : Monday, September 16, 2019 - 11:45:49 AM
Long-term archiving on : Friday, December 16, 2016 - 4:23:54 AM

bredin12eccvws.pdf

Publisher files allowed on an open archive

Fusion of Speech, Faces and Text for Person Identification in TV Broadcast

Files

Identifiers

Collections

Citation

Export

Metrics

1542

1321

Contact

Fusion of Speech, Faces and Text for Person Identification in TV Broadcast

Files

Identifiers

Collections

Citation

Export

Share

Metrics

1542

1321

Contact