A new study performed by a group of researchers has pointed out the way a new technique of artificial intelligence can be used to detect, describe, and count automatically the animals residing within their habitats. This cutting-edge technology is known as deep learning.
As a part of this technique, pictures are collected automatically by using motion-sensor cameras. These pictures can then be described automatically with the help of “deep neural networks.” With this system, animal identification can be automated by near about 99.3 percent. The accuracy rate is around 96.6 percent, which is similar to that of the estimations made physically by human volunteers.
The study’s lead author, Jeff Clune from the University of Wyoming, said in a statement, “This technology lets us accurately, unobtrusively and inexpensively collect wildlife data, which could help catalyze the transformation of many fields of ecology, wildlife biology, zoology, conservation biology and animal behavior into ‘big data’ sciences.” Further, the research manager of the Uber’s Artificial Intelligence Labs, added, “This will dramatically improve our ability to both study and conserve wildlife and precious ecosystems.”
The “deep neural networks” are basically a type of “computational intelligence” that works by analyzing how the brains of animals work when they see the world. The neural networks require large amounts of precisely labeled data to work correctly.
The first author of the study, Sadegh Norouzzadeh said, “Deep learning is still improving rapidly, and we expect that its performance will only get better in the coming years. Here, we wanted to demonstrate the value of the technology to the wildlife ecology community, but we expect that as more people research how to improve deep learning for this application and publish their datasets, the sky’s the limit. It is exciting to think of all the different ways this technology can help with our important scientific and conservation missions.”
The finding of the study was featured in the Proceedings of the National Academy of Sciences (PNAS). The study paper is written by Jeff Clune; Mohammad Sadegh Norouzzadeh, Clune’s Ph.D. student; Anh Nguyen, an ex Ph.D. student of Clune; Margaret Kosmala at the Harvard University; Meredith Palmer and Craig Packer at the University of Minnesota; and Ali Swanson at the University of Oxford.