Retrieving Diverse Social Images at MediaEval 2017: Challenges, Dataset and Evaluation

By Maia Zaharieva, Bogdan Ionescu, Alexandru Lucian Ginsca, Rodrigo L.T. Santos, and Henning Müller

Abstract

This paper provides an overview of the Retrieving Diverse Social Images task that is organized as part of the MediaEval 2017 Benchmarking Initiative for Multimedia Evaluation. The task addresses the challenge of visual diversi cation of image retrieval results, where images, metadata, user tagging pro les, and content and text models are available for processing. We present the task challenges, the employed dataset and ground truth information, the required runs, and the considered evaluation metrics.

Reference

M. Zaharieva, B. Ionescu, A. Ginsca, R. Santos, H. Müller: "Retrieving Diverse Social Images at MediaEval 2017: Challenges, Dataset and Evaluation"; Talk: MediaEval Benchmarking Initiative for Multimedia Evaluation, Dublin, Ireland; 09-13-2017 - 09-15-2017; in: "MediaEval 2017 Multimedia Benchmark Workshop", CEUR-WS, Vol-1984 (2017), ISSN: 1613-0073; 3 pages.

BibTeX

@inproceedings{TUW-262849,     author = {Zaharieva, Maia and Ionescu, Bogdan and Ginsca, Alexandru Lucian and Santos, Rodrigo L.T. and M{\"u}ller, Henning},     title = {Retrieving Diverse Social Images at MediaEval 2017: Challenges, Dataset and Evaluation},     booktitle = {MediaEval 2017 Multimedia Benchmark Workshop},     year = {2017},     journal = {CEUR-WS},     volume = {Vol-1984},     numpages = {3},     url = {http://ceur-ws.org/Vol-1984/Mediaeval_2017_paper_1.pdf},     issn = {1613-0073},     keywords = {image diversification, benchmark, dataset, MediaEval},     note = {talk: MediaEval Benchmarking Initiative for Multimedia Evaluation, Dublin, Ireland; 2017-09-13 -- 2017-09-15} } Click into the text area and press Ctrl+A/Ctrl+C or ⌘+A/⌘+C to copy the BibTeX into your clipboard… or download the BibTeX.