The goal of Big Ocean, Big Data was to establish FathomNet, a new baseline dataset optimized to directly accelerate development of modern, intelligent, automated analysis of underwater visual data.

Ocean-going platforms and instruments are integrating high-resolution camera feeds for observation and navigation, producing a deluge of visual data that rapidly outpaces researchers’ abilities to process and analyze them. Recent advances in machine learning enable fast, sophisticated analysis of visual data, but have had limited success in the ocean due to lack of data set standardization and insufficient availability of existing, expertly curated imagery. Publicly released in 2021, FathomNet addresses this need by aggregating images from multiple sources to create a publicly available, expertly curated underwater image training database.

Team: Kakani Katija, Eric Orenstein, Brian Schlining, Lonny Lundsten, Kevin Barnard, Giovanna Sainz, Océane Boulais, Alexandra Lapides, Benjamin Woodward, Katy Croff Bell

Funded by: National Geographic Society and NOAA Office of Ocean Exploration & Research. Additional funding has been raised from numerous sources to continue work beyond the scope of the initial Rapid Field Deployment, including the Packard Foundation and NSF.

2021 status: Continuing