Afeka College Seeks to Identify COVID-19 Carriers Using Voice Analysis 

Article Featured Image

Afeka Tel-Aviv Academic College of Engineering has begun an extensive research project, led by the Israeli Ministry of Defense (MOD), to detect potential carriers of COVID-19 using voice and speech processing. The study is being conducted by the Afeka Center for Language Processing (ACLP) with Rabin Medical Center, Sheba Hospita, several high-tech companies, and various MOD units.

"The target of the project is to develop an innovative algorithm that will be a breakthrough in the identification of potential carriers of the COVID-19 virus," said Ami Moyal, a speech processing expert and president of Afeka College of Engineering, in a statement. "I am optimistic that this impressive and unique collaboration will yield results and contribute to public health in Israel ;and throughout the world."

Voice and speech samples (recordings) are being collected from COVID-19 patients at various stages of the disease, as well as from healthy individuals. Another control group of COVID-19-negative patients suffering from respiratory symptoms (such as flu symptoms) are also being recorded to differentiate them from COVID-19-positive patients. The goal is to create a representative database to be used in the research.

The data will be processed using a variety of tools, advanced speech processing techniques, and machine learning algorithms. In addition to studying the effect of the coronavirus on speech and the human voice, the primary purpose is to differentiate between healthy individuals and potential carriers of the COVID-19 virus.

"The Afeka Center for Language Processing has been active in speech processing research for over a decade. During this time, the center has developed tools that are being used  to diagnose diseases and assess their progress. We look forward to the ACLP knowledge and algorithms being adapted to assist in this national project," Moyal said.

SpeechTek Covers
for qualified subscribers
Subscribe Now Current Issue Past Issues