This research explores a fresh approach to the selection and weighting of classical image features for infrared object detection and target-like clutter rejection. Traditional statistical techniques are used to calculate individual features, while modern supervised machine learning techniques are used to rank-order the predictive-value of each feature. This paper describes the use of Decision Trees to determine which features have the highest value in prediction of the correct binary target/non-target class. This work is unique in that it is focused on infrared imagery and exploits interpretable machine learning techniques for the selection of hand-crafted features integrated into a pre-screening algorithm.
Companies are increasingly facing the challenges of a persistent cyber threat landscape. By means of AI, cyber attacks can be efficiently conducted more successful through offensive AI. As for cyber defense, AI can be also utilized against cyber threats (defensive AI). Due to limited resources, especially in small and medium-sized companies (SMEs), there is a need to deploy more effective defensive cyber security solutions. Precisely, the adaptation of AI-based resilient defenses must be driven forward. Therefore, the aim of this paper is to identify and evaluate AI-related use cases with a high impact potential on the cyber security level, while being applicable to SMEs at the same time. In order to reach the research goal, an extensive literature review of several online catalogs, surveys and online platforms was conducted. In conclusion, seven crucial AI-based security features were outlined that are providing a high impact potential to the security level for SMEs. Afterwards, the results are discussed and set into a broader context. Even though AI-based security solutions are providing a large range of advantages, certain challenges and barriers using AI-related security applications are addressed in the paper as well. A high need for usable state of the art AI based cyber security solution for SMEs was identified.