Israel Develops Voiceprint Recognition Against Covid-19 Virus

A few days ago, the COVID-19 virus swept the world. To deal with this infectious disease with a strong infectious capacity, the biggest problem is not the treatment, but the detection! Although the mortality rate of this disease is less than that of severe infectious diseases such as SARS, once the medical system is overwhelmed and collapses, its mortality rate will rise sharply! For this reason, all countries want to find the most convenient, fastest and sufficiently accurate detection method.
Recently, Israel, a technology powerhouse, has developed a screening method that is different from traditional body temperature and nucleic acid tests. On March 24, the Israeli Defense Ministry said it was developing a method to detect the COVID-19 virus by voice, or to identify the voice of an infected person. In this regard, some netizens said in disbelief: this kind of thing does not look unscientific at all, is it metaphysics?
The identification method is based on the rule that, because the COVID-19 virus affects the respiratory system, the patient's voice and breathing pattern will also show signs, a unique "sound fingerprint", or it can help detect whether an infection.
Israeli company Vocalis Health will collaborate with the hospital to lead the study by sampling the sounds of confirmed patients through a mobile app. Researchers will analyze speech samples based on AI algorithms to identify the characteristics of speech. The company's website allows anyone to send voice samples, hoping to use this data to prioritize testing and hospitalization.
Not only that, it has just developed a detection method called "combined detection method". Using this method, the "joint samples" of dozens of subjects can be combined for detection at one time. The system can identify individual positive samples in the "joint sample", and only in rare cases, such as when the "joint sample" is positive overall, each sample needs to be tested separately.

Post a Comment

0 Comments