This project is part of ongoing research at TNO and you will be fully integrated into our research group (Intelligent Imaging). The project will start with a deep-dive in recently published literature, to identify current methods for adversarial attacks and potential technology to detect and/or counter such attacks. Relevant datasets of military vehicles with manual annotations, both images and videos, are readily available for this project as part of our ongoing research. You will use these datasets to perform various adversarial attacks and evaluate different techniques to detect and/or counter these adversarial attacks. If time permits, we would like to study additional methods to train AI-technology to be robust against adversarial attacks, so that attack detection/countering might not even be necessary. In addition, we encourage your own initiative and are open for suggestions and new research directions related to this topic. The project will end with a written report of your recommendations (with possibility for conference/journal publication) and a presentation highlighting your results and future directions.