Few-Shot Siamese Transfer Learning for Cross-Vessel SSE
摘要
Given the complex environment of the open sea, SSE is of utmost significance for marine systems. To make safe and efficient decisions, existing environmental conditions must be appropriately perceived and processed. External sea loads are vital for the control and operation of marine systems, especially with the advancement of autonomous ships, where the need to perceive the external environment has become more pressing. Sea state information is invaluable for on-board decision-making, personnel safety, and operational efficiency. It is also critical for a wide range of marine operations, as waves significantly affect vessel safety and fuel-efficient navigation, as noted by [1]. Accurate prediction of wave-induced loads and responses requires a comprehensive understanding of on-site sea state information. With the advent of the ship intelligence era, artificial intelligence is poised to play a pivotal role in the accurate and autonomous perception of the external environment, as demonstrated by [2].