LoBIAG: A location-based collaborative image annotation game

One of the effective approaches for managing large amounts of image data is (semantic) annotation which on its own is considered as a difficult task for machines. To deal with this issue, leveraging humans' cognitive abilities has been become a popular trend within the recent years. Notwithstanding, when human annotators have no accurate contextual and/or location-related knowledge about the subject matter, the quality of annotation/labeling process will be faced with some challenges. Also, in some cases due to lack of such knowledge, the final results may be severely affected so that those will be unusable at all. In order to deal with the aforementioned issues and incorporating location-specific information in the process, a location-based game with the purpose of image annotation and based on the collaborative intelligence of human participants, entitled LoBIAG, is proposed. The rationale behind the work, its architecture, workflow and performance analysis are discussed in details in the paper.