World-Wide Scale Geotagged Image Dataset for Automatic Image Annotation and Reverse Geotagging

Hatem Mousselly-Sergieh (1) , Mario Döller , Elod Egyed-Zsigmond (1) , Daniel Watzinger (2) , Bastian Huber (2) , Harald Kosch (2)

1 DRIM - Distribution, Recherche d'Information et Mobilité
2 University of Passau

In this paper, a dataset of geotagged photos on a world-wide scale is presented. The dataset contains a sample of more than 14 Million geotagged photos crawled from Flickr with the corresponding metadata. To guarantee the spatial representativeness of the dataset, a crawling approach based on the small-world phenomena and the Flickr friendship's graph is applied. Furthermore, the noisiness of user-provided tags is reduced through an automatic tag cleaning approach. To enable efficient retrieval, photos in the dataset are indexed based on their location information using quad-tree data structure. The dataset can assists different applications, especially, search-based automatic image annotation and reverse geotagging.

Domaines: Informatique [cs]

Hatem Mousselly-Sergieh, Mario Döller, Elod Egyed-Zsigmond, Daniel Watzinger, Bastian Huber, et al.. World-Wide Scale Geotagged Image Dataset for Automatic Image Annotation and Reverse Geotagging. Multimedia Systems Conference 2014, MMSys '14, Mar 2014, Singapore, Singapore. pp.47-52, ⟨10.1145/2557642.2563673⟩. ⟨hal-01301036⟩