Marine biologists have crowd-funded a new facial recognition software that will essentially create a Whale Facebook.
This move to create a Facebook for whales comes after years of scientists not being able to quickly identify most of the remaining North Atlantic right whales. The right whale is one of the most endangered whale species, with only around 500 right whales remaining in the world.
One of the big problems scientists face while studying and trying to keep the right whale from falling to extinction is that they simply can’t easily identify the whale when it is spotted. In the past, when whales were spotted while out on an aerial survey, they usually couldn’t be identified until hours or even days later when they could be manually compared to other photos back in the lab.
One day after Christian Khan, a biologist at the National Oceanic and Atmospheric Administration’s (NOAA) Northease Fisheries Science Center was comparing some newly taken photos and searching for matches she took a break and logged into Facebook. It instantly showed her a photo of herself that someone else had taken and asked her if she would like to confirm herself.
At that moment she knew that she needed to create something similar to Facebook’s algorithms for identifying right whales. After searching through various data companies she finally took her problem to Kaggle, the self-described home of data science competitions, to find an answer.
The contest ended in January and the winning team was declared. They were able to build an algorithm that can successfully identify right whales with an 87-percent accuracy. So now it’s up to Khan to decide, along with other scientists, if they should continue with their Facebook for whales and create a program using the algorithm or keep working to try to develop something better.
Follow me on Twitter @Chrisbuck243