When you hear facial recognition, what comes to your mind? Improving biometric security for humans, right? Recently, a team of French scientists in Zimbabwe created a facial recognition technology to identify individual giraffes. The new learning computer system could help prevent giraffe extinction.
Being the tallest animals on earth, giraffes are iconic African mammals. They play an essential role in the environment where they live. However, individual giraffes’ distinct coating patterns, consisting of brown blotches on a tan background, make them difficult to differentiate. Thus, researchers are relying on facial recognition to identify them which is a crucial step in tracking and checkmating their declining population. Speaking to RFI, the lead author of the new research, Vincent Miele, from the University of Lyon’s Laboratory of Biometry and Evolutionary Biology, said
“To our knowledge, this is the first attempt in using deep learning techniques for this task.”
How the Technology Works and why it is Important
According to the researchers, the deep learning technology, commonly used in facial recognition in humans, is 90% accurate. The team trained the technology to identify individual giraffes in Hwange Park. From 2014 to 2018, they photographed about 400 giraffes in Hwange. Out of the 4,000 pictures taken, they also performed a training set by cropping the pictures to show the animals’ flanks. These scientists then inputted the photographs into the facial recognition system known as Convolutional Neural Network (CNN). According to Miele,
“The advantage of deep learning is that once the computer has been trained, it is very fast and can tackle dozens of images in a few seconds. Deep learning’s algorithms can outperform any other algorithms in terms of predicting performance.”
However, camera traps (remote devices that have sensors to capture the motion of passing animals) can generate thousands of images, and training computers to capture images is crucial for scientists. According to the French scientists, they have trained their facial recognition technology to just five photographs per animal. The captured images are amplified in the laboratory to create variability.
Future of the Facial Recognition Technology
Even though the technology sometimes creates poor-quality pictures, the scientists observed only a few incorrect matches. The co-author of the study, Christophe Bonenfan, told RFI,
“We observed very little or wrong matching. When a match is reported the result is really good and reliable.”
Besides preventing giraffe extinction, biologists and other scientists can also adopt the CNN technology. This is because the system obtains important data on group composition, life histories, or the movement of different animal species. According to Bonenfan,
“Like most giraffe populations in the world, abundance is decreasing at Hwange and obviously everybody is hunting for an explanation.”
The CNN technology was created on an existing framework that utilizes Artificial Intelligence (AI) to differentiate between giraffes. No doubt, tech innovations like this will help solve giraffe extinction and keep tabs on other animals, too. Perhaps, scientists can use the technology to identify and tag viable males for artificial birth methods.
Possible Causes of Giraffes Extinction
The decline in the population of iconic giraffes species in Africa is due to several reasons. However, the biggest threats to giraffe extinction come from habitat losses, human population growth, poaching, disease, war, and civil unrest.
The International Union For Conservation of Nature revealed nine subspecies of giraffe. Most of them are near threatened, vulnerable, or critically endangered. An independent group known as the Giraffe Conservation Foundation is currently in partnership with 16 African countries. According to the foundation, the number of giraffes in Africa decreased by 30% since the 1980s. Hopefully, the new technology will help put an end to the looming giraffe extinction.