The new AI-powered app can identify and differentiate a child’s normal voice signals from abnormal ones that indicate illness. The technology can be used both at home and in medical institutions.
Experienced parents and health care providers can fairly accurately distinguish many of the needs of children by the nature of their crying. Although each infant’s screams are unique, they do share some common features if the reasons are the same. Despite this, identifying hidden patterns in such signals is a challenge. However, a team of researchers from Northern Illinois University and the College of New Jersey has found a solution that helps to understand exactly what the baby is trying to say..
Engineers and biologists have jointly developed a special automatic speech recognition algorithm that detects the characteristics of each child’s cry. To analyze and classify these signals, the system processed huge arrays of audio data, separating the necessary elements from them against the background of the general noise..
The efficiency of the trained algorithm does not depend on the specific user or toddler. Therefore, it is versatile and analyzes various aspects of the cry, determining the cause of anxiety and urgency..
Scientists compare a child’s signals to a separate language, which contains a wealth of information about his health. Minor differences allow it to be translated. The team hopes the results of their research will help parents and healthcare providers better understand the needs of their babies, which will help them grow healthier..
Amazon also recently introduced a new AI-powered document processing product that fully automates data entry..
text: Ilya Bauer, photo: Unsplash