At dAIEDGE, we continue to highlight the people and projects driving innovation in AI at the edge through our ‘Meet the Winners’ series. In this first interview, we feature Kosta Pavlović, one of the winners of our First Open Call, and the developer of the project SEAL (Deep video watermarking for real-time deployment on the Edge).
In today’s landscape, shaped by rapid advances in generative artificial intelligence, distinguishing between real and synthetic content is becoming increasingly difficult. Kosta’s project addresses this growing challenge by focusing on video watermarking as a key technology to ensure authenticity. SEAL aims to develop lightweight, deep learning-based watermarking models capable of running directly on low-power edge devices, such as smart cameras, where content is first captured. By embedding authentication markers at the point of creation, the project contributes to building more trustworthy and reliable AI systems.
One of the main challenges SEAL tackles is the limitation of current watermarking approaches, which are often too computationally demanding for real-time deployment in resource-constrained environments. By designing efficient models adapted to these constraints, the project opens the door to scalable applications in areas such as smart cities, where ensuring the integrity of visual data is essential.
During his seven-month journey within dAIEDGE, Kosta developed his work at Università degli Studi di Modena e Reggio Emilia. We would like to express our sincere appreciation to UNIMORE for their role as a hosting institution, providing continuous support, mentorship and valuable expertise—particularly in the field of smart city solutions, a natural application domain for SEAL.
Reflecting on his experience, Kosta highlights the importance of this support and collaboration: “I’m very satisfied with the support I got from dAIEDGE and foremost from my mentors at the University of Modena and Reggio Emilia, because of their experience and projects in smart city solutions.”
Kosta’s work exemplifies the type of innovation that dAIEDGE seeks to foster: solutions that combine technical excellence with real-world applicability. By enabling efficient and trustworthy video processing at the Edge, SEAL contributes to addressing one of the key challenges of the AI ecosystem today.
This interview marks the beginning of a series in which we will continue to explore the stories, ideas and impact of our Open Call winners. See the full interview clicking here. Stay tuned for more interviews on our YouTube channel.