July 12, 2024


Source: https://www.nature.com/articles/s41598-022-19939-2.pdf

The ocean is changing at an unprecedented rate, making it challenging to maintain responsible stewardship while visually monitoring vast amounts of marine data. The amount and rate of the necessary data gathering are outpacing our capacity to process and analyze them quickly as the research community seeks baselines. The lack of data consistency, inadequate formatting, and the desire for significant, labeled datasets have all contributed to the limited success of recent advancements in machine learning, which have enabled quick and more complex visual data analysis.

In order to meet this requirement, several research institutions worked with MBARI to speed up ocean research by utilizing the capabilities of artificial intelligence and machine learning. One such outcome of this partnership is FathomNet, an open-source image database that employs cutting-edge data processing algorithms to standardize and aggregate carefully curated labeled data. The team believes that using artificial intelligence and machine learning will be the only way to speed up critical studies on ocean health and remove the bottleneck for processing underwater imagery. Details regarding the development process behind this new image database can be found in a recent research publication in Scientific Reports journal.

Machine learning has historically transformed the field of automated visual analysis, partly thanks to vast volumes of annotated data. When it comes to terrestrial applications, the benchmark datasets that machine learning and computer vision researchers swarm to are ImageNet and Microsoft COCO. To give researchers a rich, engaging standard for underwater visual analysis, the team created FathomNet. In order to establish a freely accessible, highly maintained underwater image training resource, FathomNet combines images and recordings from many different sources.

Research workers from MBARI’s Video Lab carefully annotated data representing nearly 28,000 hours of deep-sea video and more than 1 million deep-sea photos that MBARI gathered during 35 years. About 8.2 million annotations documenting observations of animals, ecosystems, and objects are present in the video library of MBARI. This comprehensive dataset serves as a priceless tool for the institute’s researchers and their international collaborations. Over 1,000 hours of video data were gathered by the Exploration Technology Lab of the National Geographic Society from various marine habitats and places across all ocean basins. These recordings have also been used in the cloud-based collaborative analysis platform developed by CVision AI and annotated by experts from the University of Hawaii and OceansTurn.

Additionally, in 2010, the National Oceanic and Atmospheric Administration (NOAA) Ocean Exploration team aboard the NOAA Ship Okeanos Explorer gathered video data using a dual remotely operated vehicle system. In order to annotate gathered videos more extensively, they started funding professional taxonomists in 2015. Initially, they crowdsourced annotations through volunteer participating scientists. A portion of MBARI’s dataset, as well as materials from National Geographic and NOAA, are all included in FathomNet.

Since FathomNet is open source, other institutions can readily contribute to it and utilize it in place of more time- and resource-consuming, conventional methods for processing and analyzing visual data. Additionally, MBARI started a pilot initiative to use machine learning models trained on data from FathomNet to analyze video taken by remotely controlled underwater vehicles (ROVs). Using AI algorithms raised the labeling rate tenfold while reducing human effort by 81 percent. Machine-learning algorithms based on FathomNet data may revolutionize ocean exploration and monitoring. One such example includes using robotic vehicles equipped with cameras and enhanced machine learning algorithms for automatic search and monitoring of marine life and other underwater things.

With ongoing contributions, FathomNet currently has 84,454 images that reflect 175,875 localizations from 81 different collections for 2,243 concepts. The dataset will soon have more than 200 million observations after obtaining 1,000 independent observations for more than 200,000 animal species in various positions and imaging settings. Four years ago, the lack of annotated photos prevented machine learning from examining thousands of hours of ocean film. By unlocking discoveries and enabling tools that explorers, scientists, and the general public may utilize to quicken the pace of ocean research, FathomNet, however, turns this vision into a reality.

FathomNet is a fantastic illustration of how collaboration and community science may promote innovations in our understanding of the ocean. The team believes the dataset can aid in accelerating ocean research when understanding the ocean is more crucial than ever, using data from MBARI and the other collaborators as the foundation. The researchers also emphasize their desire for FathomNet to function as a community where ocean aficionados and explorers from all walks of life may share their knowledge and skills. This will act as a springboard to address problems with ocean visual data that otherwise would not have been achievable without widespread participation. In order to speed up the processing of visual data and create a sustainable and healthy ocean, FathomNet is constantly being improved to include more labeled data from the community.

This Article is written as a research summary article by Marktechpost Staff based on the research paper 'FathomNet: A global image
database for enabling artifcial intelligence in the ocean'. All Credit For This Research Goes To Researchers on This Project. Check out the paper, tool and reference article.
Please Don't Forget To Join Our ML Subreddit

Khushboo Gupta is a consulting intern at MarktechPost. She is currently pursuing her B.Tech from the Indian Institute of Technology(IIT), Goa. She is passionate about the fields of Machine Learning, Natural Language Processing and Web Development. She enjoys learning more about the technical field by participating in several challenges.




Source link