AI for Marine Perception

Program Description

Robust knowledge is urgently needed to support sustainable use and restoration of marine  ecosystems. However, the dynamic nature of marine environments presents distinct challenges  for eƯective monitoring and data-driven decision-making. Shifting environmental conditions,  limited light penetration, and complex sound propagation constrain the direct application of  conventional sensing and machine learning techniques developed for terrestrial use.  Consequently, research in marine environments demands tailored AI approaches that integrate  domain expertise, multimodal sensing, and adaptive learning strategies.  

The P1 Program, AI for Marine Perception, seeks to catalyze interdisciplinary collaboration between researchers  and practitioners working at the intersection of AI, marine science, and underwater robotics, with the aim to strengthen the scientific foundation for applying AI in aquatic environments.
This will be achieved by identification of shared research challenges, community building,  knowledge exchange, and creating opportunities for collaborative innovation. The vision is to  promote dialogue between technical and domain experts, fostering new ideas and cross-sector  collaborations that can advance the understanding and management of our marine  environments. The program will facilitate national and international workshops serving as  platforms for disseminating recent advances, discussing data collection and evaluation  practices, and exploring potential joint initiatives. Representative topics include multimodal  learning from visual, acoustic, environmental, and remote data; adaptive autonomy for marine  robotics; AI for spatial perception and geometric scene understanding; and AI-driven  approaches for environmental and infrastructure monitoring and mapping.