Active Selection with Label Propagation for Minimizing Human Effort in Speaker Annotation of TV Shows

Budnik Mateusz 1, 2 Johann Poignant 1 Laurent Besacier 1, 2 Georges Quénot 3
3 MRIM - Modélisation et Recherche d’Information Multimédia [Grenoble]
LIG - Laboratoire d'Informatique de Grenoble, Inria - Institut National de Recherche en Informatique et en Automatique
Abstract : In this paper an approach minimizing the human involvement in the manual annotation of speakers is presented. At each iter- ation a selection strategy choses the most suitable speech track for manual annotation, which is then associated with all the tracks in the cluster that contains it. The study makes use of a system that propagates the speaker track labels. This is done using a agglomerative clustering with constraints. Several dif- ferent unsupervised active learning selection strategies are eval- uated. Additionally, the presented approach can be used to ef- ficiently generate sets of speech tracks for training biometric models. In this case both the length of the speech track for a given person and its purity are taken into consideration. To evaluate the system the REPERE video corpus was used. Along with the speech tracks extracted from the videos, the op- tical character recognition system was adapted to extract names of potential speakers. This was then used as the 'cold start' for the selection method.
Document type :
Conference papers
Liste complète des métadonnées

Cited literature [18 references]  Display  Hide  Download

https://hal.archives-ouvertes.fr/hal-01055704
Contributor : Laurent Besacier <>
Submitted on : Wednesday, August 13, 2014 - 12:32:53 PM
Last modification on : Monday, February 11, 2019 - 4:36:02 PM
Document(s) archivé(s) le : Thursday, November 27, 2014 - 12:27:58 AM

File

slam2014_submission_11.pdf
Files produced by the author(s)

Identifiers

  • HAL Id : hal-01055704, version 1

Citation

Budnik Mateusz, Johann Poignant, Laurent Besacier, Georges Quénot. Active Selection with Label Propagation for Minimizing Human Effort in Speaker Annotation of TV Shows. Workshop on Speech, Language and Audio in Multimedia (SLAM 2014), Sep 2014, Penang, Malaysia. 5 p. ⟨hal-01055704⟩

Share

Metrics

Record views

317

Files downloads

262