
MBARI and WoRMS: Collaborating to build open datasets of ocean life with @FathomNet marinespecies.org/news.php?p… #marinespecies @LifeWatchVLIZ @UNOceanDecade

An open, expertly curated underwater image training set that will allow us to better understand the ocean and its inhabitants.

MBARI and WoRMS: Collaborating to build open datasets of ocean life with @FathomNet marinespecies.org/news.php?p… #marinespecies @LifeWatchVLIZ @UNOceanDecade

Incredibly excited about this new collaboration between @FathomNet/@oceanvisionai/@MBARI_News and @WRMarineSpecies, working together to enhance public biodiversity datasets! #onwardsanddownwards
MBARI and WoRMS: Collaborating to build open datasets of ocean life with @FathomNet marinespecies.org/news.php?p… #marinespecies @LifeWatchVLIZ @UNOceanDecade

Are you ready for a #TechTuesday challenge? Join the @FathomNet #Kaggle competition with @fgvcworkshop to show us how you would develop a model that can recognize 18 supercategories and flag unknown objects. The deadline to submit is June 12. ➡️ bit.ly/3wLZXkM #CVPR2024

Are you ready for a challenge? We just launched a competition on @kaggle for the @fgvcworkshop at @CVPR. The challenge is to develop a model that can recognize the original 18 supercategories and flag unknown objects. More info🧵👇

Developing novel or general category discovery for ocean research will help scientists to rapidly find unlabeled objects in historic datasets and enable creative field campaigns to seek out new life. Join the competition: kaggle.com/competitions/fath…

How do you get a machine learning model to flag animals that are new and different? Join our @kaggle competition with @fgvcworkshop to show us how YOU would develop a model that can recognize 18 supercategories and flag unknown objects. kaggle.com/competitions/fath… #CVPR2024

Developing novel or general category discovery for ocean research will help scientists to rapidly find unlabeled objects in historic datasets and enable creative field campaigns to seek out new life. Join the competition: kaggle.com/competitions/fath…

The test set contains these same 18 supercategories AND new ones not represented in the training data. The challenge is to develop a model that can recognize the original 18 supercategories and flag unknown objects.

For this competition, we have curated data from the broader FathomNet image set that is emblematic of this challenging use-case. The training set contains 18 supercategories of marine animals collected along the ocean floor.

This combination of dense, image-based sampling and a poorly defined space of classes makes for a challenging computer vision problem: how do you get a machine learning model to flag animals that are new and different?

The ocean is one of the least-explored habitats on Earth. It is home to diverse animals, many that are either unknown to science or poorly understood. Scientists are increasingly leveraging ocean-going camera systems to monitor populations and discover new animals.

Thank you to all who participated in the 2024 FathomNet Workshop. Recordings are now available on YouTube. youtube.com/playlist?list=PL…

Join us on March 14th for the virtual FathomNet Workshop! We'll be sharing an exclusive preview of #FathomGPT, our latest project, and we're eager to gather valuable feedback from our community. Register now and mark your calendars. oceanvisionai.org/news/fatho…

📅 March 14 Session 1 @ 0800-1100 PT Session 2 @ 1600-1900 PT 📍 Zoom 🔗 oceanvisionai.org/news/fatho…

REGISTER NOW! At the third #FathomNet Workshop we will share progress and exciting updates about FathomNet with special sessions for programmers and marine scientists looking to utilize our open-source image and machine learning model repository.

📢JOB ALERT! Work with us and the @OceanVisionAI team as an #ML/#MLOps Researcher! We just opened up this #ResearchEngineer position at @MBARI_News and we're hoping to review applications before the end of the month. Please share! mbari.org/job-opening/ovaire…