Artificial intelligence (AI) is the ability of a computer or a robot to do tasks that are usually done by humans. A few decades ago, it may have seemed like a sci-fi movie, but these days, we have manufacturing robots, virtual travel booking agents, self-driving cars, and AFID. AFID (Automated Fish Identification platform) is a system that has been created to identify, count, and measure fish (//youtu.be/S_S-16j0bZg) from underwater imagery. It uses machine learning and AI to automatically gather information about fish, making data gathering more efficient, accurate, and cheaper.
How does it work?
- A baited remote underwater video system is set up to capture footage of the fish on a reef.
- The footage from the BRUV is run through the AFID platform.
- Fish ecologists watch the video footage that AFID processes and makes corrections, filling in any missing data.
- AFID then re-trains and relearns to become more accurate over time.
What’s the point?
Monitoring fish biodiversity and biomass are the best indicators of ecological health. Thousands of hours of BRUV footage are collected annually, but the time it takes to analyze it all means that there is a serious lag time between data collection and actual results being delivered to managers and policy makers. Speeding up this process allows policy decisions on fish stocks and quotas, environmental impact assessment, and ecological protection to be better informed.
You can be a part of this!
AFID is currently in the proof of concept stage of the BRII – AIMS challenge. If you have data or time to contribute to this project please take a look at the website (//www.afid.io/home) or contact Dan Marrable.
For more BRUV footage: