Key points are not available for this paper at this time.
Fiber-Optic Hydrophones (FOHs) are starting to gain interest for Passive Acoustic Monitoring (PAM) applications. One of the fiber optic sensing technologies that has been shown to have practical applicability in harsh environments is the use of sensors based on Fiber Bragg Grating (FBG). Through the construction of a prototype and its experimental validation, it has been demonstrated that a 7 Multi-Core Fiber (MCF) adhered longitudinally to a membrane on the top of a hollow cylinder transforms acoustic pressure waves into mechanical deformations detectable by the 7 FBGs. The combination of this structure with a Deep Convolutional Neural Network based on a U-Net architecture allows to reduce the FOH noise and increase its sensitivity offering a reference framework for the design of FBG-based hydrophones in specific applications. • A 7-core Fiber Bragg Grating hydrophone enhances acoustic sensitivity. • U-Net convolutional networks effectively denoise multi-core spectrograms. • Training on modeled fin whale calls transfers successfully to real recordings. • Demonstrates a lightweight, non-electrical alternative for acoustic monitoring.
Miralles et al. (Tue,) studied this question.