Between December 15-18, 2024, the IEEE Big Data Conference took place in Washington, DC, USA, gathering researchers, innovators, and practitioners worldwide. Among the highlights was the IBDA4FCT workshop, a premier platform addressing the transformative role of Big Data in combating crime and terrorism, supporting Law Enforcement Agencies (LEAs), and fostering advanced border management techniques. This workshop also featured insights backed by the Horizon Europe-funded Ceasefire project (GA. 101073876), underscoring collaborative efforts to empower LEAs with cutting-edge AI tools and protocols to address cross-border illicit activities effectively.
As part of this esteemed forum, the Cygnus project team proudly presented our paper S03221, titled: "Adversarial Explanations for Informed Civilian and Environmental Protection". Authored by Theodora Anastasiou, Ioannis Pastellas, and Sophia Karagiorgou, this work delves into:
🛡️Enhancing AI transparency and robustness to build trust in decision-making for critical infrastructures.
🛡️Addressing vulnerabilities like adversarial attacks on AI systems.
🛡️Efficient processing of multimodal Big Data from IoT and sensory devices.
🛡️Tackling the complexity of AI adoption in sensitive civilian and environmental applications. Our approach introduces and benchmarks adversarial explanations—an innovative method for ensuring robust, interpretable AI models capable of withstanding unseen perturbations and attacks.
📑 For more details, you can read the full paper and explore our findings in depth.
🌟 The Cygnus team remains committed to advancing AI technologies that prioritize reliability, robustness, and transparency in tackling some of society's most pressing challenges.