top of page

Accelerating ocean discovery with artificial intelligence

VIDEO: MBARI

MISSION

FathomNet streamlines access to artificial intelligence and machine learning tools to accelerate the analysis of ocean visual data.

SOLUTIONS

Three interconnected software platforms to make AI and ocean research more accessible and impactful

Portal

The FathomNet Portal is a globally accessible end-to-end tool that interfaces with FathomNet Database and FathomVerse. It utilizes machine learning to dramatically accelerate the analysis of ocean imagery.

Database

The FathomNet Database is a growing database of expert-annotated images and machine learning models that provides machine learning-ready data representing all underwater life for software engineers and marine scientists.

FathomVerse

FathomVerse is an app that enables users to interact with real ocean imagery, helping to train AI models through community science.

Visit FathomVerse

LATEST NEWS

FathomNet

Join New FathomNet Kaggle Competition

FathomNet

FathomNet at OSM 2026

FathomNet

Open Data Opens New Opportunities

COMMUNITY

@kakanikatija delivered a captivating presentation at @yachtclubmonaco during @the_explorers_club Day of Exploration!

Bioengineer Kakani Katija founded the groundbreaking Bioinspiration Lab at the Monterey Bay Aquarium Research Institute (@MBARI_News) to investigate how ocean animals can inspire new technologies and reveal the mysteries of the deep sea. Leveraging cutting-edge robotics and novel AI solutions, she brings the laboratory to the ocean by developing innovative imaging technologies. Through the FathomNet program, Kakani connects experts in engineering, computer science, ocean exploration, and community engagement with a common goal to increase our capacity to measure and monitor biodiversity in a changing ocean.

Join the FathomNetwork to get involved (link in bio).

@kakanikatija delivered a captivating presentation at @yachtclubmonaco during @the_explorers_club Day of Exploration!

Bioengineer Kakani Katija founded the groundbreaking Bioinspiration Lab at the Monterey Bay Aquarium Research Institute (@MBARI_News) to investigate how ocean animals can inspire new technologies and reveal the mysteries of the deep sea. Leveraging cutting-edge robotics and novel AI solutions, she brings the laboratory to the ocean by developing innovative imaging technologies. Through the FathomNet program, Kakani connects experts in engineering, computer science, ocean exploration, and community engagement with a common goal to increase our capacity to measure and monitor biodiversity in a changing ocean.

Join the FathomNetwork to get involved (link in bio).

We have a major problem. Many of the annotated images in the #FathomNetDatabase have sparse coverage–only 1 or some of the animals are labeled. We’re asking the computer vision community for help to develop better object detectors.

If the training data contains sparsely labeled images, can you build object detection systems that identify all animals in the test set?

Join our Kaggle challenge w/ #LifeCLEF & #FGVC13 to show us your solution! LINK IN BIO

#FathomNet #FathomNetCLEF #ComputerVision #CV4Ecology #objectdetection #Kaggle #machinelearning #datascience

We have a major problem. Many of the annotated images in the #FathomNetDatabase have sparse coverage–only 1 or some of the animals are labeled. We’re asking the computer vision community for help to develop better object detectors.

If the training data contains sparsely labeled images, can you build object detection systems that identify all animals in the test set?

Join our Kaggle challenge w/ #LifeCLEF & #FGVC13 to show us your solution! LINK IN BIO

#FathomNet #FathomNetCLEF #ComputerVision #CV4Ecology #objectdetection #Kaggle #machinelearning #datascience

Are you ready for a challenge?

For this year’s FathomNet Kaggle Competition, we’re tackling one of the biggest challenges in computer vision for marine imagery: incomplete labels. Marine biologists are experts in specific animals so they only label what they know in each image. Consequently, annotators may carefully focus on jellyfish in a given image, while ignoring the dozen other creatures swimming by. Just because something isn't labeled doesn't mean it isn't there, and that's a serious problem for anyone trying to teach a computer to recognize all ocean life.

The FathomNet 2026 challenge addresses this real-world constraint by focusing on positive-unlabeled learning for object detection. Developing robust methods in this setting will reduce reliance on expert annotations, accelerate dataset curation, and broaden the usability of large-scale archives like the FathomNet Database.

Dive in today! Link in bio.

This competition is held jointly as part of: The LifeCLEF 2026 Lab of the CLEF 2026 conference, and of the FGVC13 Workshop organized in conjunction with the CVPR 2026 conference.

*Correction: the source for the image used in the background is NOAA Ocean Exploration

#Kaggle #DataScience #AI #ComputerVision #MachineLearning #FathomNetCLEF #CVPR #LifeCLEF #FathomNetDatabase #FGVC13

Are you ready for a challenge?

For this year’s FathomNet Kaggle Competition, we’re tackling one of the biggest challenges in computer vision for marine imagery: incomplete labels. Marine biologists are experts in specific animals so they only label what they know in each image. Consequently, annotators may carefully focus on jellyfish in a given image, while ignoring the dozen other creatures swimming by. Just because something isn't labeled doesn't mean it isn't there, and that's a serious problem for anyone trying to teach a computer to recognize all ocean life.

The FathomNet 2026 challenge addresses this real-world constraint by focusing on positive-unlabeled learning for object detection. Developing robust methods in this setting will reduce reliance on expert annotations, accelerate dataset curation, and broaden the usability of large-scale archives like the FathomNet Database.

Dive in today! Link in bio.

This competition is held jointly as part of: The LifeCLEF 2026 Lab of the CLEF 2026 conference, and of the FGVC13 Workshop organized in conjunction with the CVPR 2026 conference.

*Correction: the source for the image used in the background is NOAA Ocean Exploration

#Kaggle #DataScience #AI #ComputerVision #MachineLearning #FathomNetCLEF #CVPR #LifeCLEF #FathomNetDatabase #FGVC13

Attending #OSM26 in Glasgow? Join us for the FathomNet Town Hall on Thursday, Feb. 26 at 12:45. We'll demo the FathomNet Database, Portal, and FathomVerse—plus share how you can contribute visual data, analyze ocean video, and expand engagement through gaming.

Come with questions, leave with solutions. See you there! Link in bio to save to your calendar.

Attending #OSM26 in Glasgow? Join us for the FathomNet Town Hall on Thursday, Feb. 26 at 12:45. We'll demo the FathomNet Database, Portal, and FathomVerse—plus share how you can contribute visual data, analyze ocean video, and expand engagement through gaming.

Come with questions, leave with solutions. See you there! Link in bio to save to your calendar.

Annotated images from the FathomNet Database were used to train and evaluate the new Segment Anything Model 3 (SAM 3), adding complex underwater scenes to the broad training dataset and expanding the variety of environments the model can recognize. Two outputs from this process—the set of 280, 118 segmentation masks used for training and the set of 14,247 segmentation masks used for testing—are now available to download from the FathomNet Database.

The addition of Meta’s segmentation masks to the FathomNet Database represents a significant step forward in bridging the gap between AI research and ocean science. We look forward to continuing to expand the FathomNet Database to help scientists everywhere study, understand, and protect the ocean.

Link in bio to learn more!

#FathomNet #SegmentAnything

@mbari_news @meta

Annotated images from the FathomNet Database were used to train and evaluate the new Segment Anything Model 3 (SAM 3), adding complex underwater scenes to the broad training dataset and expanding the variety of environments the model can recognize. Two outputs from this process—the set of 280, 118 segmentation masks used for training and the set of 14,247 segmentation masks used for testing—are now available to download from the FathomNet Database.

The addition of Meta’s segmentation masks to the FathomNet Database represents a significant step forward in bridging the gap between AI research and ocean science. We look forward to continuing to expand the FathomNet Database to help scientists everywhere study, understand, and protect the ocean.

Link in bio to learn more!

#FathomNet #SegmentAnything

@mbari_news @meta

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! Link in bio!

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! Link in bio!

You're invited to our upcoming #FathomNetWorkshop on Wednesday, April 9, 2025! To accommodate different time zones, we’re hosting two sessions:
🕗 Session 1: 0800-1000 Pacific
🕓 Session 2: 1600-1800 Pacific

What’s on the Agenda?
✅ FathomNet Updates – Database news, Kaggle competition, WoRMS collaboration, and a sneak peek at our 2025 plans.
✅ FathomVerse Insights – What we’ve learned from the launch plus new features in the 2.0 update!
✅ How-To Sessions – Get hands-on guidance on contributing data, verifying entries, using FathomNet models, and engaging enthusiasts with FathomVerse.

Link in bio to register!

📸 @mbari_news

You're invited to our upcoming #FathomNetWorkshop on Wednesday, April 9, 2025! To accommodate different time zones, we’re hosting two sessions:
🕗 Session 1: 0800-1000 Pacific
🕓 Session 2: 1600-1800 Pacific

What’s on the Agenda?
✅ FathomNet Updates – Database news, Kaggle competition, WoRMS collaboration, and a sneak peek at our 2025 plans.
✅ FathomVerse Insights – What we’ve learned from the launch plus new features in the 2.0 update!
✅ How-To Sessions – Get hands-on guidance on contributing data, verifying entries, using FathomNet models, and engaging enthusiasts with FathomVerse.

Link in bio to register!

📸 @mbari_news

Big news from FathomNet! 🎉🌊

A recent video brought 2,800+ new players to #FathomVerse in one day! Now, we’re launching FathomVerse 2.0 in March, kicking off our third Kaggle competition, and hosting the 4th Annual #FathomNet Workshop on April 9th.

Get all the details and more in our Winter 2025 newsletter! Link in bio!

📸 @mbari_news

Big news from FathomNet! 🎉🌊

A recent video brought 2,800+ new players to #FathomVerse in one day! Now, we’re launching FathomVerse 2.0 in March, kicking off our third Kaggle competition, and hosting the 4th Annual #FathomNet Workshop on April 9th.

Get all the details and more in our Winter 2025 newsletter! Link in bio!

📸 @mbari_news

Are you ready for a challenge?

Calling all data scientists and machine learning engineers! Join the FathomNet Kaggle Competition with FGVC Workshop to show us how you would develop a model that can accurately classify varying taxonomic ranks. Developing these solutions for ocean research will enable scientists to process and explore ocean data more efficiently.

Learn more at the link in bio!

📸 @mbari_news

#CVPR2025 #FathomNet #MachineLearning #DataScience #KaggleCompetition

Are you ready for a challenge?

Calling all data scientists and machine learning engineers! Join the FathomNet Kaggle Competition with FGVC Workshop to show us how you would develop a model that can accurately classify varying taxonomic ranks. Developing these solutions for ocean research will enable scientists to process and explore ocean data more efficiently.

Learn more at the link in bio!

📸 @mbari_news

#CVPR2025 #FathomNet #MachineLearning #DataScience #KaggleCompetition

.pro-gallery-wix-wrapper {display: block !important;} .pro-gallery-wix-wrapper .gallery-item-container {opacity: 1 !important; display: block !important;}

Join our network of marine scientists, programmers, and enthusiasts all working to understand life in the ocean.

Follow FathomNet

Contact

fathomnet@mbari.org

RESOURCES

Lightbulb and book icon in blue on a pale background with abstract grid and ocean waves, symbolizing ideas and knowledge.

Lightbulb and book icon in blue on a pale background with abstract grid and ocean waves, symbolizing ideas and knowledge.

Identifying Ocean Animals with FathomVerse

Resources for Educators

Starfish labeled "Asteroidea" on a monitor, set against a textured underwater background. The mood is scientific and exploratory.

Starfish labeled "Asteroidea" on a monitor, set against a textured underwater background. The mood is scientific and exploratory.

The Art of Annotation: How to design effective guides for computer vision annotation tasks

Portal Resources

Close-up of a red jellyfish, Tiburonia granrojo, floating in dark water.

Close-up of a red jellyfish, Tiburonia granrojo, floating in dark water.

How to submit localized image annotations to the FathomNet Database

Database Resources

bottom of page