A citizen science and machine learning approach to identify fish in baited underwater videos.
School of Blue maomao in Poor Knights Islands Marine Reserve.
Photo by the New Zealand Department of Conservation
A biodiversity hotspot
Counting how many fish are in the ocean and how different threats like fishing, pollution and climate change influence marine ecosystems is a great challenge.
Baited underwater video
A camera pointing to a bait is an ideal non-invasive method to monitor fish populations.
Biologists use this method to know more about the abundance of carnivorous and scavenger fishes.
Hours and hours of footage
Baited underwater videos provide crucial information on the populations of fish species. However, the manual classification of the footage is time consuming.
Wildlife.ai is developing a citizen science and machine learning approach to classify target fish species from underwater footage recorded around Aotearoa/New Zealand.
Using AI to Identify New Zealand Fish – Project Completion Update
In this post, we explain the different machine learning experiments we performed to create models to automatically [...]
Spyfish Aotearoa: using AI to identify New Zealand fish
In this post, we will share an update on how we are using machine learning to identify [...]
How we used ML to identify fish
In this short post, I will share how our team of volunteers (data rangers) used machine learning [...]
Spyfish Aotearoa – Open for contributions
The community science page of Spyfish Aotearoa is active and +700 underwater videos have already been classified! [...]
Spot the fish!
The first version of Spyfish Aotearoa is online, join us and help identify New Zealand fish in our [...]