Multimodal AI-enhanced ship detection for mapping fishing vessels and informing on suspicious activities

Alessandro Galdelli, Gagan Narang, Rocco Pietrini, Micol Zazzarini, Andrea Fiorani, Anna Nora Tassetti

Abstract: Highlights•Novel method tackles suspicious maritime activities using remote sensing imagery.•Multimodal framework integrates Sentinel-1 SAR, Sentinel-2, and optical imagery.•Enhanced YOLO model detects small ships in cluttered maritime environments.•Custom HS3-S2 dataset improves detection accuracy across multiple metrics.•Eigen-CAM ensures model transparency; CodeCarbon is used to report CO2<math><msub is="true"><mrow is="true"></mrow><mrow is="true"><mn is="true">2</mn></mrow></msub></math> emissions during training.