For a parent, their child’s cry can be worrisome and infuriating at the same time. A researching team based in the United States has established a novel Artificial Intelligence technique which is able to recognize and then differentiate between normal and abnormal bouts of crying in babies. This AI method is able to check whether the crying is owed to hunger, ailment or anything else. The IEEE/CAA Journal of AutmaticaSinica printed this newest study.
Even though every infant’s cries are distinct, all cries have some mutual characteristics if they are a consequence of the same cause. The novel study utilizes a particular algorithm that identifies the various sorts of baby cries, based on automatic recognition of speech.
The research group utilized compressed sensing in order to analyze a large amount of data effectively, so they could evaluate and categorize the received signals. Compressed sensing is a technique that rebuilds a signal based on a small amount of data collected. This is particularly for sound distinguishing in noisy areas, which is usually where a baby cries.
The researchers, for their investigation, generated a new code for the recognition of crying, which is capable of differentiating the reasons behind normal crying and abnormal crying both, in loud surroundings.
The code is independent of the type of crying registered. This means it can be utilized in practical situations so as to identify and categorize the different characteristics of crying and better comprehend the reasons behind a baby’s cries and the urgency behind the cries.
Co-author Lichuan Liu stated how, like in a different language, different crying sounds hide behind them loads of health-associated data. The data is carried in the distinctions between the various sound signals.
She added how their ultimate aim is healthier children and a lesser amount of stress on parents.