Abstract

Images play a crucial role in online news perception. Images catch users? attention and strongly affect how they interpret the news. Images serve different roles, e.g., visualizing a scene discussed in the text, highlight certain aspects by a stock photo, or showing archived footage of a relevant person or organization. News Images as part of MediaEval 2022 aims to gain more insight into the interplay of images and texts in different news domains. Participant access a large set of articles and accompanying images collected from general online news portals and an RSS-based news stream. In contrast to NewsImages 2021 data come from different news sources. Thus, this year?s task facilitates comparing image usage on different portals and analyzing transfer learning strategies. This paper describes the NewsImages task, explains the dataset and evaluation metrics. It draws connections to existing research.

@inproceedings{NewsImagesInMediaEval2022,
	author    = {Benjamin Kille and Andreas Lommatzsch and {\"O}zlem {\"O}zg{\"o}bek and Mehdi Elahi and Duc-Tien Dang-Nguyen},
	title     = {News Images in MediaEval 2022},
	booktitle = {Proceedings of the MediaEval Benchmarking Initiative for Multimedia Evaluation 2022},
	year      = {2023},
	issn      = {1613-0073},
	url       = {https://ceur-ws.org/Vol-3583/},
	note      = {urn:nbn:de:0074-3583-1},
	location  = {Bergen, Norway},
	publisher = {CEUR Workshop Proceedings}
}