MediaEval Benchmark: Social Event Detection in collaborative photo collections

ABSTRACT In this paper, we present an approach to detect social events in collaboratively annotated photo collections as part of the MediaEval Benchmark. We combine various information from tagged photos with external data sources to train a classification model. Experiments based on the MediaEval Social Event Detection Dataset demonstrate the effectiveness of our approach. Categories and Subject Descriptors H.3.1 [ Information Storage and Retrieval ]: Content Analysis and Indexing; H.3.3 [ Information Systems ]: Information Storage and Retrieval General Terms Design, Experimentation, Performance Keywords Benchmark, Photo Collections, Classification, Event Detection 1. INTRODUCTION The Internet enables people to host, access and share their photos online, e.g. through websites like Flickr. Collaborative annotations and tags are commonplace on such services. The information people assign vary greatly, but often seem to include some kind of references to