FEDERATED LEARNING FOR SHIP DETECTION IN REMOTE SENSING IMAGERY: CURRENT PROGRESS AND FUTURE DIRECTIONS FEDERATED LEARNING FOR SHIP DETECTION IN REMOTE SENSING IMAGERY: CURRENT PROGRESS AND FUTURE DIRECTIONS

Ship detection plays a vital role in maritime surveillance by enabling vessel traffic monitoring, maritime security, the prevention of illegal fishing, and search-and-rescue operations. Advances in deep learning, transformer-based architectures, and multimodal remote sensing have substantially improved ship detection performance in both optical and synthetic aperture radar (SAR) imagery. Nevertheless, practical maritime monitoring systems often rely on distributed data sources managed by various organizations, complicating centralized data sharing due to privacy, security, and regulatory constraints. Federated Learning (FL) has emerged as a promising approach that facilitates collaborative model training without the need to exchange raw data. This article provides a comprehensive review of ship detection and federated learning within the context of privacy-preserving maritime surveillance. The evolution of ship detection technologies, key federated learning algorithms, and recent applications of federated learning in remote sensing are systematically examined. In addition, current challenges such as data heterogeneity, communication overhead, computational complexity, and limited benchmark availability are discussed. Potential research directions for developing scalable, privacy-preserving, and distributed ship detection systems are also identified. This review serves as a structured reference for researchers and practitioners seeking to integrate federated learning with maritime remote sensing applications. Ship detection plays a vital role in maritime surveillance by enabling vessel traffic monitoring, maritime security, the prevention of illegal fishing, and search-and-rescue operations. Advances in deep learning, transformer-based architectures, and multimodal remote sensing have substantially improved ship detection performance in both optical and synthetic aperture radar (SAR) imagery. Nevertheless, practical maritime monitoring systems often rely on distributed data sources managed by various organizations, complicating centralized data sharing due to privacy, security, and regulatory constraints. Federated Learning (FL) has emerged as a promising approach that facilitates collaborative model training without the need to exchange raw data. This article provides a comprehensive review of ship detection and federated learning within the context of privacy-preserving maritime surveillance. The evolution of ship detection technologies, key federated learning algorithms, and recent applications of federated learning in remote sensing are systematically examined. In addition, current challenges such as data heterogeneity, communication overhead, computational complexity, and limited benchmark availability are discussed. Potential research directions for developing scalable, privacy-preserving, and distributed ship detection systems are also identified. This review serves as a structured reference for researchers and practitioners seeking to integrate federated learning with maritime remote sensing applications.