STORM graffiti/tagging detection dataset

This object detection dataset was created by University of West Attica, by collecting images containing at least one graffiti/tagging, using the camera of a smartphone and exploiting the STORM crowdsensing application. STORM (Safeguarding Cultural Heritage through Technical and Organisational Resources Management) is a HORIZON 2020 funded European Union Cultural Heritage project that aims at the protection of Cultural Heritage through a combination of technical and organizational resources (http://www.storm-project.eu). The included images display structures, such as walls, pavements, benches and statues that are located in the municipality of Athens and its suburbs. The dataset consists of 1022 images with graffiti/tagging annotations and are split into 80% (813 images) for training and validation and 20% (209 images) for testing.