NSF Convergence Accelerator Track E: Ocean Vision AI: Scaling up Visual Observations of Life in the Ocean Using Artificial Intelligence

点击翻译按钮获取中文摘要

OIA - 2137977 NSF Convergence Accelerator Track E: Ocean Vision AI: Scaling up visual observations of life in the ocean using artificial intelligenceAbstractThis project will scale up society’s observational capabilities to fully explore the ocean and discover the full spectrum of animals that live there. The ocean represents the largest habitable ecosystem on the planet, yet less than 5% of that volume has been explored, and nearly 50% of marine life are yet to be described. To close this gap, underwater imaging, a major sensing modality for marine biology, is being deployed on a diverse array of platforms. However, as more visual data are collected, the community faces a data analysis backlog. This project, Ocean Vision AI, will deploy artificial intelligence (AI) and machine learning to address this backlog by automating underwater image and video analysis. The research activities include contributions from the research community as well as the general public, video game players, and advanced high school and community college students. Together, Ocean Vision AI will be used to directly accelerate the automated analysis of underwater visual data to enable scientists, explorers, policymakers, storytellers, and the public, to learn, understand, and care more about the life that inhabits the oceans.The research team will engage researchers and innovators across numerous sectors (e.g., academic, government, non-profit, for-profit) to advance society’s observational capabilities of marine life. Ocean Vision AI (artificial intelligence) aims to provide a central hub for incubating groups conducting research that use imaging, AI, open data, and hardware/software; create data pipelines from existing image and video data repositories; and provide project tools for coordination. In addition, Ocean Vision AI will leverage public participation and engagement via game development, and will result in data products that are shared with researchers as well as other US and global open data repositories. This research will facilitate large-scale, spatiotemporal surveys of underwater communities, geomorphology, and marine debris, and accelerate the discovery of marine life by making well classified images widely available to experts using taxonomic metadata standards. Moreover, this project will be able to coordinate and scale underwater computer vision research and machine learning algorithm development via novel training data through the FathomNet database. This project will ultimately enable innovative intellectual pursuits in fields as diverse as marine biology, fisheries, biological oceanography, underwater optics and computer vision, artificial intelligence, ocean engineering, biomechanics, environmental biology, human-computer interaction, game-based education, and community contributions to science.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.