Retrieving diverse social images at mediaeval 2016: Challenge, dataset and evaluation

B. Ionescu (1) , A.L. Gînscǎ (2) , M. Zaharieva (3) , B. Boteanu (1) , M. Lupu (3) , H. Müller (4)

1 LAPI - Laboratoire d'Analyse et Traitement d'Images
2 DIASI (CEA, LIST) - Département Intelligence Ambiante et Systèmes Interactifs
3 TU Wien - Vienna University of Technology = Technische Universität Wien
4 HES-SO Valais-Wallis - Haute École Spécialisée de Suisse Occidentale Valais-Wallis

This paper provides an overview of the Retrieving Diverse Social Images task that is organized as part of the Media-Eval 2016 Benchmarking Initiative for Multimedia Evaluation. The task addresses the problem of result diversification in the context of social photo retrieval where images, metadata, text information, user tagging profiles and content and text models are available for processing. We present the task challenges, the proposed data set and ground truth, the required participant runs and the evaluation metrics.

Mots clés: Data set; Evaluation metrics; Ground truth; nocv1; Photo retrieval; Social images; Text information; Text models

Domaines: Informatique [cs]

B. Ionescu, A.L. Gînscǎ, M. Zaharieva, B. Boteanu, M. Lupu, et al.. Retrieving diverse social images at mediaeval 2016: Challenge, dataset and evaluation. 016 Multimedia Benchmark Workshop, MediaEval 2016, 2016, Hilversum, Netherlands. ⟨cea-01843182⟩