Sea-Scan: High-Accuracy, ML-based Dark Vessel Detection and Localisation via Weakly Supervised DAS Monitoring

2026-06-19

AI Summary

This study addresses the challenge of detecting non-cooperative maritime vessels—so-called “dark ships”—which lack reliable labels for supervised learning. To overcome this limitation, the authors propose a weakly supervised learning approach for vessel detection and localization using distributed acoustic sensing (DAS) data. This work represents the first application of weak supervision to DAS, leveraging noisy and incomplete Automatic Identification System (AIS) signals as proxy labels to train the model without requiring precise annotations. Experimental results demonstrate that the proposed method achieves a detection rate of 97.8% with a false alarm rate of only 1.98%, even under high label noise conditions, thereby significantly enhancing the capability to monitor non-cooperative maritime targets.

Abstract

We present an ML-based vessel detection and localization system, trained with weak supervision from imperfect AIS labels, that achieves a 97.8% detection rate at 1.98% false-trigger rate, successfully identifies dark-vessel events from unlabeled data.

Problem

• dark vessel detection
• weak supervision
• AIS
• maritime surveillance
• vessel localization

Innovation

• weakly supervised learning
• dark vessel detection
• AIS-based monitoring
• machine learning for maritime surveillance
• anomaly localization