What do artificial intelligence, marine life, taxonomy, and images have in common? The answer in one word is FathomNet, but the collaboration goes way beyond that. MBARI and WoRMS team up to enhance the use of AI in marine research.
The Monterey Bay Aquarium Research Institute (MBARI, California) is combining advances in marine science and engineering to understand our changing ocean. FathomNet is a global, open-source image and machine learning model repository founded by MBARI Principal Engineer Kakani Katija. The goal of FathomNet is to enable artificial intelligence to better understand our ocean and its inhabitants. But before AI can be used to analyze ocean visual data, it must be trained to accurately identify marine life. FathomNet contains over 100.000 labeled images, and shares machine learning models that can be used to identify ocean animals, making it a one-of-a-kind resource and tool within the combined world of AI and marine biodiversity research.
Recently, WoRMS and MBARI have signed a Memorandum of Understanding to enable cooperation on the labelling of data curated from public datasets. This data will then be ingested into FathomNet and WoRMS, and will further stimulate the use of biodiversity information systems for scientific research.
FathomNet is part of a larger program called Ocean Vision AI, which is endorsed by the Ocean Decade, that is creating interconnected software solutions tools to accelerate ocean exploration and discovery. Ocean Vision AI is a research and development collaboration led by MBARI, CeNCOOS, CVision AI, Ocean Discovery League, and Purdue University. This innovative program seeks to build a global community of programmers, marine scientists, and enthusiasts working toward understanding life in the ocean. The Ocean Vision AI team recently released FathomVerse, a mobile game where players can interact with real ocean images to improve the AI that helps researchers study marine life.
FathomVerse was inspired by other community science initiatives such as iNaturalist, but goes further into the blue, inviting all ages to dive into ocean exploration and discovery. This collaboration between MBARI and WoRMS strengthens outreach in several fields: engineering, science, and the public.
This agreement also contributes to the continued development of the LifeWatch Species Information Backbone (LW-SpIBB), which also aims to bring together taxonomic and species-related data, thereby filling the gaps in our knowledge and reducing duplication of efforts. The WoRMS Data Management Team (DMT) at VLIZ is supported by LifeWatch Belgium, part of the E-Science European LifeWatch Infrastructure for Biodiversity and Ecosystem Research. LifeWatch is a distributed virtual laboratory, used for different aspects of biodiversity research. The LW-SpIBB aims to bring together taxonomic and species-related data to fill the gaps in our knowledge.