Over the past three decades, tens of thousands of hours of deep ocean video and still imagery have been captured by the oceanographic community at considerable effort and expense, but only a small percentage of that video has been viewed and analyzed. With such a small percentage of the overall ocean visually explored, we can’t afford to leave any possible discoveries buried in the data.
Ocean AI proposes to dramatically accelerate the process of viewing, analyzing, and initially classifying those vast video archives using machine learning, making the discoveries within available to the entire scientific community and the public.
Our goal is to build a globally accessible platform where hundreds of thousands of hours of deep ocean video can be automatically transcoded, objects detected, and initial classification done with minimal human intervention. As the product is trained and informed at scale, it can grow into a platform that can be introduced into the ocean video collection workflow in real time, dramatically increasing the value of ocean observations for the entire scientific community.
An easy-to-use web-based experience where various users can:
Learn about the product, team, and mission
Upload/process various quantities of deep-sea video (all types and formats)
Participate as an expert or non-expert model trainer
Analyze the video for detection and classification of various marine species
Search for, and be alerted to, findings of interest in analyzed videos
Export that analysis in some usable form for their research
Develop a comprehensive, open source, ML model library for all deep ocean species and tools to continually train and inform these models (both automatically and manually).
Develop a team to support all aspects of the process from intake, analysis, and output and to begin to grow this product into a future platform.