AI identification of birds will ‘revolutionise’ conservation
BIRDS could be spared the stressful process of identification tagging after researchers developed facial recognition software which can recognise individual birds, something humans are unable to do.
The artificial intelligence tool has an accuracy of up to 92 per cent and could boost conservation efforts by “revolutionising” the identification process.
Individual recognition of animals is one of the most expensive and timeconsuming problems in research, and techniques such as putting colour bands on birds’ legs can sometimes cause them anxiety and stress. But these issues could be solved thanks to a technique known as deep learning, which specialises in classifying images.
Companies like Facebook have access to millions of pictures of people that are voluntarily tagged by users. But acquiring such labelled photographs of animals is difficult, and has led to a bottleneck in research. Scientists overcame this by building feeders with camera traps and sensors.
The study, published in Methods in Ecology and Evolution, involved collecting thousands of labelled images of birds. They included wild great tits, sociable weavers and captive zebra finches – among the most studied animals in nature. They then trained a computer model to recognise the individual birds. The programme successfully identified over 90 per cent of great tits and sociable weavers and 87 per cent of captive zebra finches.
Dr Ferreira, of the University of Montpellier, France, and lead author said: “Deep learning has the potential to revolutionise the way researchers identify individuals. To our knowledge, this is the first successful attempt at performing such an individual recognition in small birds.”
Bird populations around the world face a number of existing pressures including climate change, intensive farming and deforestation.
Global warming has already had a significant impact on numbers – increasing the extinction risk for many species. Dr Ferreira added: “The development of methods for automatic, noninvasive identification represents a major breakthrough in research.”