What question did this study set out to answer?

This research aims to develop a deep learning-based detector for identifying previously unseen vessel classes in underwater acoustic settings.

May 14, 2026

Unseen class detection in underwater acoustic target recognition

Key Points

This research aims to develop a deep learning-based detector for identifying previously unseen vessel classes in underwater acoustic settings.
Utilized a deep learning framework to distinguish novel vessel classes from known classes.
Collected training and testing data from the same measurement site, with unseen classes in the test samples.
Evaluated multiple detection techniques to assess their effectiveness under shared environmental conditions.
Specific detection techniques demonstrated effectiveness in recognizing unseen vessel classes while operating in similar acoustic environments.
Potential implications for enhanced maritime surveillance and naval operations were identified.

Abstract

This study provides deep learning-based detector which recognizes previously unseen vessel classes for passive sonar system. We focus on a scenario in which training and testing data are collected from the same measurement site, but the test samples include vessel class excluded during training. To address this challenge, we devise specific detection framework capable of distinguishing novel class based on acoustic characteristics learned from known classes. This study aims to evaluate multiple specific detection techniques for their effectiveness in separating seen and unseen class under shared environmental conditions. This approach has significant potential for maritime surveillance, environmental monitoring, and naval operations. This work was supported by the Korea Research Institute for Defense Technology Planning and Advancement (KRIT) through a grant funded by the Korean government (DAPA, Defense Acquisition Program Administration) under Grant No. KRIT-CT-23-026, as part of the project “Integrated Underwater Surveillance Research Center for Adapting Future Technologies” (2024), and also by the Ministry of Oceans and Fisheries, Korea, through the project titled “Fostering Talent in Advanced Ship Blue Tech” (Project No. Rs-2025-02221147).

Mark Helpful

Bookmark

Relay