Research towards AI technologies that facilitate everyday life has strongly emerged during the last years in order to modernize operations and optimize productivity in a wide range of sectors and applications. AI research in defense is revolutionizing military operations, enhancing decision-making, surveillance, and autonomous systems. However, when referring to the defense domain, AI technologies insert additional challenges, as security matters are under serious consideration to avoid any undesirable disclosure of sensitive information. Ethical concerns, cybersecurity risks, and the need for robust, explainable AI models pose significant hurdle towards AI implementation in the sector.

The ΑΙ specificities of the Defense Domain that M4D group addresses involve:
• Data frugality.
• Cybersecurity.
• Adaptation in new environments.
• Threats associated with CBRN hazards.

M4D group combines research from various domains and techniques in order to adapt and reuse already existing developments that can also satisfy the needs of the sector. More specifically, M4D performs research activities towards the development of frugal models that can be trained with the use of less or no annotated data via different learning paradigms, in order to tackle the scarcity or unsuitability of training data. Advanced machine and deep learning methods are being developed to detect chemical or biological warfare agents, with the scope of tackling CBRN threats. Leveraging from the group’s extensive background in fusion techniques, additional knowledge is extracted from outcomes of different modalities in a complementary manner, to provide improved situational awareness and decision support schemas to the operators.