FathomNet

Have you played FathomVerse?? If you're not already hooked, you should give it a try because FathomVerse is transforming the future of community science and advancing ocean exploration through innovative collaboration. FathomVerse gamifies ocean exploration to train AI and discover ocean life. The FathomVerse team created a mobile game to enlist the help of ocean enthusiasts and speed up breakthroughs in ocean science. By participating, players contribute directly to cutting-edge ocean research, accelerating AI-assisted image analysis to improve our understanding of marine ecosystems. #FathomVerse #OceanGames #OceanConservation #BlueEconomy #SustainableSolutions MBARI Schmidt Ocean Institute Schmidt Marine Technology Partners Monterey Bay Aquarium NOAA Ocean Exploration Ocean Networks Canada Ocean Census Ocean Discovery League Dalio Philanthropies OceanX

#ICYMI In case you missed it, the summer newsletter from FathomNet has all the latest updates from our busy program! Check it out ➡️ https://lnkd.in/gGrwvMpw

As we enter a new season, we're excited to share updates from the #FathomNet program. This summer, we celebrated the completion of our third and biggest Kaggle competition with 109 participants and 5 winning teams. We also successfully connected #FathomVerse with the #FathomNetDatabase, a key milestone in developing interconnected solutions to enhance AI for ocean discovery. FathomVerse has now contributed 99,124 labeled images, representing 47% of the total dataset. We appreciate your continued support as we work to connect communities and accelerate ocean discovery. Read our latest newsletter for more updates from our team! 🔗 https://lnkd.in/gGrwvMpw MBARI FGVC Workshop National Science Foundation (NSF) Kakani Katija Monterey Bay Aquarium NOAA Ocean Exploration Schmidt Ocean Institute Schmidt Marine Technology Partners Dalio Philanthropies Laura Valentine Chrobak

Great to be on this Playing for the Planet panel at #JacksonWild with this all star cast. The main take-aways from the discussion were: 1. Games and nature-based documentaries can work together; 2. Players are leaning in on these themes; 3. The behaviour change opportunity is significant. Thanks to Adrienne Hall from Sound Off, Alan Gershenfeld from E-Line Media Joost Daniels from MBARI and Samantha-Lynn Martinez from Jackson Wild for joining the conversation and Jane Kachmer & Tria Thalman for offering up the opportunity.

Putting ocean AI to the test 🌊 🤖 🌊 MBARI researchers are leveraging the power of artificial intelligence and machine learning to accelerate efforts to study the ocean, but they need tools to evaluate their effectiveness. MBARI’s Video Lab has developed the first publicly available benchmark dataset for multi-object tracking in deep-sea video. Led by Summer Intern Elaine Liu from the University of Virginia, DeepSea MOT tests the performance of AI models in real ocean environments. The dataset features seafloor and midwater scenes, with “simple” and “difficult” modes covering 2,400 total frames and more than 57,000 annotated objects, to determine how well object detection and tracking models perform under real deep-sea challenges—low visibility, a bustling community environment crowded with animals, transparent organisms, camera motion, and more. MBARI is committed to sharing our data to advance marine science and engineering. The tools, codes, and workflows available on GitHub (https://lnkd.in/gGNUJN3V) will provide the necessary resources to help scientists worldwide analyze underwater imagery and accelerate exploration of a rapidly changing ocean. Learn more: https://lnkd.in/gVgbm9ff

Registration is still open! Join FathomNet Engagement Coordinator Lilli Wakinekona Carlsen for this interactive online workshop! http://bit.ly/4ljxs18

Scientists are collecting massive amounts of images and video to study marine life and assess ocean health. For nearly two decades, supervised machine learning models have helped researchers analyze this deluge of visual data more efficiently. Before they can be used for ocean exploration, these machine learning models need to be trained to identify objects and animals. In a recent study, the team at FathomNet leveraged Meta’s Segment Anything Model to enrich ocean image annotations based on existing labels. Their workflow yielded promising results, especially for modernizing existing data repositories. It also suggests that future human annotation efforts could use foundation models to speed progress toward a sufficient training set to tackle domain-specific problems. Read more: https://lnkd.in/geP4Etxz

We're excited to announce that the FathomNet Database has been selected as one of 10 new datasets added to the National Artificial Intelligence Research Resource Pilot. The #FathomNetDatabase is an open-source library for machine learning models and expertly labeled ocean imagery that can help us better understand marine life and ocean health. We look forward to seeing how this integration can contribute to advances in machine learning and accelerate the analysis of ocean visual data. The FathomNet Database continues to grow with data contributed by MBARI, Schmidt Ocean Institute, NOAA Ocean Exploration, Ocean Exploration Trust, Ocean Networks Canada, Ocean Observatories Initiative & more partners. Learn more in the latest announcement from National Science Foundation (NSF): https://lnkd.in/gectXXmg

We are thrilled to share that National Geographic Magazine’s September issue features Katy Croff Bell, Ocean Discovery League, and the development of DORIS, a low-cost deep-sea imaging system created with Blue Robotics. With the ability to reach 98% of the ocean floor for a target price of under $10,000, DORIS is a breakthrough in expanding and accelerating ocean exploration. #NationalGeographic #PerpetualPlanet #deepseascience #lowcostsensors Read the full story: https://lnkd.in/eu7pDS_6 Purchase the issue: https://lnkd.in/eE_Fu3Ed Learn more about DORIS: https://lnkd.in/eDCKDwpv

Congratulations to the winning teams from FathomNet's 2025 Kaggle Competition! Several winning teams discovered that looking at the bigger picture improved their results. “Notably, we extended beyond simply analyzing cropped organisms by modeling the relationship between species and their environment, and incorporating hierarchical taxonomy into our model. We hope our approach contributes to future research in fine-grained marine species classification,” said winning team leader Donghwan Lee. Read more in a new blog by Laura Valentine Chrobak: https://lnkd.in/g8RpK8fu