Application of Image Captioning and Retrieval to Support Military Decision Making

Abstract

Image data forms an important percentage in the makeup for providing common operational pictures (COPs) required in military decision making and tactical planning, upon which critical decisions and actions are then taken. However, the considerable volume and velocity of this data places a need to process it in a timely manner with suitable meta-data (features) extracted to support information retrieval and decision making processes. In this paper we apply and explore the benefits of using image captioning whose main task is to create machine generated sentences that can describe an image appropriately. In this way user queries can be expressed using natural language sentences (scene descriptions), in order to frame and subsequently improve the efficiency with which relevant images are retrieved