Scientists from Michigan State University have developed a new software with a face recognition system that will help protect endangered primate species.
The new development, which was named PrimNet, uses the technologies of neural networks and artificial intelligence.
In addition to the high accuracy of the results compared to other analogs, the PrimNet recognition system is more economical, and also safer for tracking animals. Traditional instruments of this kind are very expensive, the prices for them can vary from 400 to 4000 US dollars. In this case, the whole process of catching and marking animals can cause them stress and damage, which is completely excluded when using the new technology.
In addition to PrimNet, an application for PrimID smartphones was created. The application contains a large database of images of primates: 280 individuals of 14 species. With its help, researchers can photograph the primates, and identify them with an accuracy of 90 percent. For example, in the identification of lemurs, the program showed an amazing accuracy of 93.75%. If the application can not find an exact match, then it will offer the 5 most possible options.
In the future, the project developers plan to expand databases of primates, improve the recognition system, as well as share the results of their activities, laying them out in open access on the Internet.