Covid Cough Identification using Machine Learning and Deep Learning Networks

• Published in: 2023 3rd International Conference on Intelligent Technologies (CONIT) pp. 1 - 4
• Main Authors: Vinod, Anand, Mohan, Neethu, S, Sachin Kumar, K P, Soman
• Format: Conference Proceeding
• Language: English
• Published: IEEE 23-06-2023
• Subjects: Cough Audio; COVID-19; Deep learning; Feature extraction; Human factors; Machine Learning; Measurement; Pandemics; Social factors
• DOI: 10.1109/CONIT59222.2023.10205895

The COVID-19 pandemic engulfed the entire world. RT-PCR assessment nowadays is a metric golden standard for contemplating COVID. However, this method takes time and violates social distancing and thus necessary steps have to be taken for its prediction in the future due to its communicable property. Cough is an important bio-marker and it contains micro patterns and audio fingerprints that can be used for classification purposes. Thus in this paper, the main focus is given to the cough data for the classification of COVID-19 diseases. The necessary features such as ZCR, MFCC, chroma STFT, roll-off, spectral centroid, and spectral bandwidth are extracted from the cough data. Traditional machine learning models as well as various deep learning models are implemented in order to find out the model with maximum accuracy for COVID-19 identification.