🤖 AI Summary
Illegal, unreported, and unregulated (IUU) fishing poses severe threats to marine ecosystems. To address this challenge, we propose an end-to-end, AI-driven global real-time monitoring framework that fuses multi-source satellite data—including Sentinel-1/2 SAR and optical imagery, Landsat 8–9, and NOAA series observations. We introduce the first set of four domain-specific computer vision models explicitly designed for heterogeneous remote sensing modalities (SAR and optical). Our framework features a novel lightweight training paradigm and edge-optimized inference architecture, enabling scalable, low-latency global deployment. All models are open-sourced and integrated into the public-good platform Skylight, where the system is already operational. Experimental results demonstrate substantial improvements in both detection accuracy and response latency for IUU fishing vessels. This work establishes a scalable, reproducible, and publicly deployable technical paradigm for marine conservation and maritime domain awareness.
📝 Abstract
Illegal, unreported, and unregulated (IUU) fishing poses a global threat to ocean habitats. Publicly available satellite data offered by NASA, the European Space Agency (ESA), and the U.S. Geological Survey (USGS), provide an opportunity to actively monitor this activity. Effectively leveraging satellite data for maritime conservation requires highly reliable machine learning models operating globally with minimal latency. This paper introduces four specialized computer vision models designed for a variety of sensors including Sentinel-1 (synthetic aperture radar), Sentinel-2 (optical imagery), Landsat 8-9 (optical imagery), and Suomi-NPP/NOAA-20/NOAA-21 (nighttime lights). It also presents best practices for developing and deploying global-scale real-time satellite based computer vision. All of the models are open sourced under permissive licenses. These models have all been deployed in Skylight, a real-time maritime monitoring platform, which is provided at no cost to users worldwide.
