The scope for implementation of artificial intelligence is increasing day by day. The result is, scientists have developed an AI system that is capable to automatically identify and count wild animals in their natural habitats.
This AI system uses motion-sensor camera that automatically collects animal's photographs and deep neural network to describe those animals. This system can automate animal identification up to 99.3% of the images. Researchers believe that this AI system can collect wildlife data in very inexpensive manner and has the potential to transform the field of ecology, wild life biology, animal behavior etc. The functioning of AI system requires huge amount of labelled data.
It is being widely used to track record of wild life in their habitats all over the world. A recent study published by computer scientist Mohammed Sadegh describes that, the researcher has tested new AI system embedded with deep neural network could identify and count species and capable to determine animal's age or classify their behavior.
They found that this AI system is capable of using 3.2 million images from the Serengeti dataset and required to label these pictures, which the system did in one hour with 96.6% accuracy and capable to save 17000 hours of human.
Margaret Kosmala from Harvard University said that, this AI system can tell you how many animals are present and from which species they belong to. It can also tell us is animals are eating, sleeping.
Collecting data automatically, accurately and in inexpensive way can transform many field. Scientist also estimates that these data can catalyze fields like zoology, conservation biology and animal behavior into data science.
Improve Wild life Conservation with AI
Beside tracking the endangered animal, Scientist have developed a method for applying morphometrics to animal footprints with an effort to expand this skill set to scale wild life conservation effort across the globe. They also told that how morphometrics can automate and scale the approach to monitor endangered wild life using SAS platform by analyzing where animals have walked.
the sophisticated algorithm not only predicts where the poachers will be in the whole area but also suggest rangers the most effective way to thwart them. Now the researchers are developing their algorithm from predictive to prescriptive analytics. They are developing AI for wild life protection.