IRIM at TRECVID 2012: Semantic Indexing and Instance Search

Nicolas Ballas 1 Benjamin Labbé 1 Aymen Shabou 1 Hervé Le Borgne 1 Philippe-Henri Gosselin 2 Miriam Redi 3 Bernard Merialdo 3 Hervé Jégou 4 Jonathan Delhumeau 4 Rémi Vieux 5 Boris Mansencal 5 Jenny Benois-Pineau 5 Stéphane Ayache 6 Abdelkader Hamadi 7 Bahjat Safadi 7 Franck Thollard 8 Nadia Derbas 7 Georges Quénot 8, * Hervé Bredin 9 Matthieu Cord 10 Boyang Gao 11 Chao Zhu 11 Yuxing Tang 11 Emmanuel Dellandrea 11 Charles-Edmond Bichot 11 Liming Chen 11 Alexandre Benoit 12 Patrick Lambert 12 Tiberius Strat 12 Joseph Razik 13 Sébastien Paris 13 Hervé Glotin 13, 14 Tran Ngoc Trung 15, 16 Dijana Petrovska-Delacrétaz 15, 16 Gérard Chollet 17 Andrei Stoian 18 Michel Crucianu 18
Abstract : The IRIM group is a consortium of French teams work- ing on Multimedia Indexing and Retrieval. This paper describes its participation to the TRECVID 2012 se- mantic indexing and instance search tasks. For the semantic indexing task, our approach uses a six-stages processing pipelines for computing scores for the likeli- hood of a video shot to contain a target concept. These scores are then used for producing a ranked list of im- ages or shots that are the most likely to contain the tar- get concept. The pipeline is composed of the following steps: descriptor extraction, descriptor optimization, classi cation, fusion of descriptor variants, higher-level fusion, and re-ranking. We evaluated a number of dif- ferent descriptors and tried di erent fusion strategies. The best IRIM run has a Mean Inferred Average Pre- cision of 0.2378, which ranked us 4th out of 16 partici- pants. For the instance search task, our approach uses two steps. First individual methods of participants are used to compute similrity between an example image of in- stance and keyframes of a video clip. Then a two-step fusion method is used to combine these individual re- sults and obtain a score for the likelihood of an instance to appear in a video clip. These scores are used to ob- tain a ranked list of clips the most likely to contain the queried instance. The best IRIM run has a MAP of 0.1192, which ranked us 29th on 79 fully automatic runs.
Type de document :
Communication dans un congrès
TRECVID - TREC Video Retrieval Evaluation workshop, Nov 2012, Gaithersburg, MD, United States. 12p., 2012
Liste complète des métadonnées


https://hal.archives-ouvertes.fr/hal-00770258
Contributeur : Georges Quénot <>
Soumis le : vendredi 4 janvier 2013 - 19:15:03
Dernière modification le : jeudi 9 février 2017 - 15:20:12
Document(s) archivé(s) le : vendredi 5 avril 2013 - 05:59:16

Fichier

Ballas-al_TRECVID2012.pdf
Fichiers produits par l'(les) auteur(s)

Identifiants

  • HAL Id : hal-00770258, version 1

Citation

Nicolas Ballas, Benjamin Labbé, Aymen Shabou, Hervé Le Borgne, Philippe-Henri Gosselin, et al.. IRIM at TRECVID 2012: Semantic Indexing and Instance Search. TRECVID - TREC Video Retrieval Evaluation workshop, Nov 2012, Gaithersburg, MD, United States. 12p., 2012. <hal-00770258>

Partager

Métriques

Consultations de
la notice

1327

Téléchargements du document

413