FathomVerse is an app that aims to grow a library of labeled imagery that can train artificial intelligence to identify marine life.

Created by MBARI, the Monterey Bay Aquarium, and our collaborators in the FathomNet Program, FathomVerse allows everyone with a smartphone or tablet to participate in ocean exploration and discovery. 

Join our community of FathomNauts—download FathomVerse today!

Download on the App Store
Download on Google Play

New Filter mini-game in FathomVerse 3.0 update

With a new mini-game and more educational content, FathomVerse 3.0 expands the app’s offerings and lets users participate in the entire image annotation process from start to finish.

A person seated in a blue armchair plays a mini-game in the FathomVerse app on her iPhone. She is wearing a gray sweater and blue jeans. On her phone screen is a grid of six photos on a blue background with app icons. She is tapping one of the photos to add a red box. In the background are a brown leather foot rest, a brown wooden table, a green houseplant in a white pot, a blue couch, and gray carpet.
The FathomVerse app’s new Filter mini-game lets users sort and match images for focused, faster labeling. Image: Calista Kerba © 2026 MBARI

New Features in FathomVerse 3.0 include: 

  • Filter Mini-game: Sort and match images for focused, faster labeling
  • Denizen Mode: Experienced FathomNauts can verify community consensus labels to help our experts confirm consensus and share more data to the FathomNet Database
  • Training Mode: Training for new mission certification is now available in Identify and Illuminate mini-games
  • Skip in Bound: Choose the skip button in Bound to move onto the next image
  • Expanded mission briefings: View more example images in the mission briefings

Download the latest version of FathomVerse on the App Store or Google Play.

Scientists are collecting massive amounts of images and video to study marine life and assess ocean health. Artificial intelligence (AI) can help researchers analyze this deluge of visual data more efficiently. However, before AI can be used to analyze the footage, it must be trained to identify specific ocean animals. The FathomVerse app aims to improve the tools needed to teach AI to study the ocean. FathomVerse enlists ocean enthusiasts worldwide to work alongside scientists to help review and label images so AI can accurately recognize ocean animals.

Inspired by other community science apps like iNaturalist and eBird, FathomVerse taps into widespread enthusiasm for ocean animals, merging science with an innovative app experience and inviting a broad community of people to participate in ocean exploration and discovery.

Five vertical screenshots illustrate gameplay in the FathomVerse app. The first screenshot on the left includes the game’s blue home screen with the FathomVerse logo and several buttons players can tap to play the game and access FathomVerse content. The second screenshot shows the library of animal training missions with circular, white cartoon animal icons. The third screenshot shows gameplay with a petal-like avatar moving through a stream of pink dots with a directional arrow, animal photos in large bubbles, and icons for game progress on a dark-blue background. The fourth screenshot shows tagging of an image in the game with a photo of an orange sea star and five circular, white cartoon icons to tag possible animals on a dark-blue background. The fifth screenshot on the right shows player progress with circular numbered white icons for game level and square photos of award videos along a vertical white dotted line on a blue background.
After completing training missions that teach users how to identify groups of marine animals, players can tag real images contributed by oceanographic research organizations and individuals. These labeled images will help machine learning models distinguish groups of marine animals so AI can accelerate analysis of ocean imagery. Image: © 2025 MBARI
Two gamers share their discoveries in the FathomVerse mobile game. The gamer in the foreground has long black hair and is wearing a tan sweater. In her left hand is an iPhone displaying a photo of a white octopus. She has her back to the camera and is pointing to the other player’s iPhone with her right hand. The player in the background has shoulder-length blonde hair and is wearing a pale-purple shirt, a white tank top, and blue jeans. She is sitting on a navy-blue couch and holding an iPhone displaying a photo of a bright-orange jelly. Large windows overlooking a sandy beach are in the background.
FathomVerse is building a community of ocean enthusiasts. Image: Marike Pinsonneault © 2025 MBARI

Collaboration is the only way to truly maximize the potential of science and technology to help us understand the ocean. MBARI is lending our expertise and resources to support much-needed efforts in developing AI tools to accelerate ocean visual data analysis. FathomVerse is one of three solutions created by the FathomNet Program, an innovative, collaborative project led by MBARI Principal Engineer Kakani Katija that seeks to make AI and ocean research more accessible and impactful.

Diverse perspectives and expertise can spark new innovations in ocean science and engineering. FathomVerse has engaged more than 30,000 passionate players, or “FathomNauts,” from more than 170 countries. Together, they have logged more than 15 million total annotations, reaching community consensus on more than 80,000 previously unlabeled images that can now be used to train AI to identify marine life.

FathomVerse is growing a community of ocean enthusiasts on Discord. Join the conversation!

An MBARI researcher with long black hair wearing a dark-blue vest sits at a desk and looks at two computer screens displaying labeled images used to train artificial intelligence to identify deep-sea animals. The screen on the left has several small thumbnails of pink corals and a larger image of a pink coral. The screen on the right has video with several colored boxes around objects in the frame. The background is a darkened office with a brown wooden door lit from the hallway outside.
Researchers in MBARI’s Video Lab have combed through thousands of hours of deep-sea footage to identify and label animals and objects. Image: © 2022 MBARI
A machine learning model places 13 multi-colored boxes around objects in an underwater video. Each box is labeled with a number, the name of the object, and a score for certainty of identification. The underlying video is brown, muddy seafloor with an orange fish at the bottom left and several stick-like sea pens across the middle of the frame.
Manually processing images and video to identify animals and objects is very time-intensive. Artificial intelligence can accelerate this analysis. Image: © 2020 MBARI
Two people seated on a couch play the FathomVerse app on their iPhones. The person on the left is wearing a beige sweater, a fuzzy pale-pink vest, and blue jeans while holding an iPhone in a clear case. The person on the right is wearing a textured pale-pink shirt, blue jeans, and white sneakers while holding an iPhone in a pink case. Both iPhone screens are displaying an origami-like blue avatar moving through a stream of dots revealing circular photos of marine life, with three round animal icons and a progress bar at the top of the screen. Blurred in the background are a white cabinet, the black plastic leg of a chair, and gray carpet.
FathomVerse invites ocean enthusiasts to work alongside scientists to grow a library of labeled imagery that can train AI to identify marine life. Image: Marike Pinsonneault © 2025 MBARI

FathomVerse was developed with &ranj Serious Games, a Netherlands-based game design studio. 

FathomVerse features imagery contributed by the ocean exploration community:
MBARI
NOAA Ocean Exploration
Ocean Networks Canada
Ocean Observatories Initiative
Ocean Exploration Trust
Schmidt Ocean Institute
Joost Daniels
Dawn Wright

Funding for FathomVerse came from the US National Science Foundation, NOAA, the David and Lucile Packard Foundation, the National Geographic Society, and the Schmidt Marine Technology Partners, a program of the Schmidt Family Foundation.

Team

Directory

Publications

All Publications
Sorry, no results were found.

Related News

All News

Technologies

All Technologies

Data

All Data