ABSTRACT
Mel frequency cepstral coefficients are one of the most prominent sets of primary features of an audio signal which are used for speech detection and cough analysis. This paper presents a new method that can overcome some of the common problems faced by using MFCCs for cough detection. In the proposed method, the most prominent part of the cough sample (HCP) is extracted and used to obtain the MFCC vectors of that particular window. These HCP MFCC vectors work as a standard comparison index for all cough samples to detect any respiratory disorders. The evaluation of the proposed method is done using 40 samples of COVID-19 patients of which 20 are positive and 20 are negative. The accuracy of the proposed method is compared with that of the standard MFCC method for the same set of samples. The proposed HCP MFCC method produces results that are 7.84% more accurate than the standard method. By bringing a standard set of comparing features that can work for almost all use cases, this method can be used as a quick identifying tool for various respiratory diseases. © 2023, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.
ABSTRACT
Mel frequency cepstral coefficients are one of the most prominent sets of primary features of an audio signal which are used for speech detection and cough analysis. This paper presents a new method that can overcome some of the common problems faced by using MFCCs for cough detection. In the proposed method, the most prominent part of the cough sample (HCP) is extracted and used to obtain the MFCC vectors of that particular window. These HCP MFCC vectors work as a standard comparison index for all cough samples to detect any respiratory disorders. The evaluation of the proposed method is done using 40 samples of COVID-19 patients of which 20 are positive and 20 are negative. The accuracy of the proposed method is compared with that of the standard MFCC method for the same set of samples. The proposed HCP MFCC method produces results that are 7.84% more accurate than the standard method. By bringing a standard set of comparing features that can work for almost all use cases, this method can be used as a quick identifying tool for various respiratory diseases. © 2023, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.
ABSTRACT
The proposed work aims to design a monitoring system that ensures COVID-free safety working environment. The proposed system has two phases. The first phase is used to monitor the vital parameters of the person entering the workplace relevant to COVID-19 test. The second phase monitors the indoor safety measures to create COVID-19-free working space. It is an inexpensive solution that aims increased COVID-19 indoor safety, with certain aspects covered like contactless sensing of temperature, heartrate monitoring, detection of mask, social distancing check, monitoring air quality, temperature, and humidity of the room. Contactless temperature sensing and pulse checking are carried out using IR sensor and heartrate sensor interfaced with Arduino Uno. The detection of mask and proper social distancing check is done by using Open CV techniques with Raspberry Pi which is equipped with a camera. The monitoring system ensures COVID-19-free safe working environment to the society. © 2022, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.
ABSTRACT
For the identification and classification of COVID-19, this research presents a three-stage ensemble boosted convolutional neural network model. A conventional segmentation model (ResUNet) is used to increase the model's performance in the initial step of processing the CXR datasets. In the second step, the CNN is used to extract the features from the pictures in the training dataset using machine learning techniques. Using machine learning (ML) techniques, the retrieved characteristics are then combined by voting in the third stage. There are 5178 aberrant CXR photos and 4310 normal CXR images used in this investigation. Models like CNN and ML can't compete with the suggested model. 99.35% of the model's measurements are accurate and precise, and 98% of its recall and F1-score are perfect. It is argued that the suggested model provides a rigorous and trustworthy evaluation of clinical decision-making in the setting of a public health crisis.