A new application based on artificial intelligence can identify and distinguish the normal voice signals of the child from anomalous, which indicate the disease. Technology can be used both at home and in medical facilities.
Experienced parents and doctors can accurately distinguish between many of the needs of children in the nature of their crying. Although the cries of each infant are unique, they have some common features if their reasons match. Despite this, the identification of hidden templates in such signals is a problem. However, a group of researchers from the University of Northern Illinois and New Jersey College found a solution that helps to understand what the baby wants to say.
Engineers and biologists jointly developed a special algorithm for automatic recognition of speech, which determines the characteristics of the cry of each child. To analyze and classify these signals, the system has handled huge arrays of sound data, highlighting the necessary elements on the background of common noise.
The efficiency of the trained algorithm does not depend on a particular user or baby. Therefore, it is universal and analyzes various aspects of the scream, determining the cause of concern and urgency.
Scientists compare the child’s signals with a separate language, which contains a lot of information about its health. Minor differences allow it to translate it. The team hopes that their research results will help parents and doctors better understand the needs of babies, which will help grow them healthier.
Recently, Amazon also introduced a new product on the basis of AI to handle documents that allow