Since the introduction of tools such as Open-AI’s Chat-GPT, Google’s Bard, and Microsoft’s new Bing, artificial intelligence has become a widely discussed topic. People are exploring the use of AI in newer and innovative ways. Reports are now suggesting that scientists can detect illness in an individual by analyzing the tone of their voice.
To teach AI to distinguish between healthy and sick individuals, researchers at the Sardar Vallabhbhai National Institute of Technology reviewed recordings of almost 630 individuals, including 111 who were experiencing cold symptoms. The researchers utilized the vocal rhythms in human speech to detect the weaker harmonics that are present in individuals who are ill. By analyzing changes in amplification, machine learning algorithms were use to identify individuals with a cold infection.
Pankaj Warule, an electronic engineer, oversaw the study that determined the differences between a “cold voice” and a healthy voice. The study found that musical instruments and human speech emit sounds at various frequencies, and an algorithm was developed that recognizes the specific frequencies that are present in individuals with a sore throat. By dissecting each participant’s speech into its spectrum of wavelengths, the researchers could determine the dominant frequency and harmonics in each case.
The study found that the technology could identify a cold in 70% of cases. With the rapid advancement of technology, it is likely that many businesses and organizations will apply this algorithm to detect employees who may be falsely claiming illness to avoid work.
You may like our posts: