A Unified Analysis of Masked Face Recognition & Social Distancing Detection

Facial recognition is an important application in today's world but given the importance of wearing face masks due to the COVID-19 pandemic, it is essential that masked face recognition performs well. It is also equally important to ensure that protective measures such as wearing face masks and social distancing are being followed. Hence, this paper introduces an approach unifying masked face recognition, mask detection, and social distancing detection into a single system. The proposed method uses RetinaFace for face detection, VGG-19 for image classification, as well as VGGFace and SVC for facial recognition. As part of experimental analysis, CNN models such as ResNet50 and Inceptionv3 were used to evaluate mask detection while FaceNet and ArcFace were used to evaluate masked face recognition. The entire system was evaluated using accuracy, precision, recall, and F1-score. The obtained results indicate that the proposed system performs efficiently. © 2022 IEEE.

Exploring SVM for Federated Machine Learning Applications

The traditional machine learning algorithms focus on centralised data repository where the aggregate data used for training is stored in a common location and processed. This approach is not suitable when data is stored in different locations and owned by different entities. Many crucial machine learning applications need computationally efficient and privacy-preserving solution. Also the central data repository has the risk of single point of failure. Federated learning is an emerging field in machine learning where the centralised concept is changed to distributed. Federated learning approach helps to train a model in machine learning without really sharing the data to a common server. In this approach, training is done locally at client side. A technique called federated averaging is applied at server side, where the model parameters from clients are aggregated and the updated parameters are computed. We propose a federated SVM architecture for solving a binary supervised classification problem. Here the experiments are done for MNIST dataset and COVID-19 dataset. Also the results are compared with centralised training approach. © 2022, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.

An unexplained death in the COVID-19 isolation ward: A case report

A 35-year-old male was referred for autopsy from a District Headquarter hospital, where he was admitted to the COVID isolation ward with suspicion of being infected. His clinical history was a day of fever with chills and abdominal pain. He was alone overnight in the isolation ward post collection of his nasal swabs for screening and blood for routine laboratory tests. However, he was found lying dead on the floor within 18 hours of hospitalization © 2021. Journal of Indian Academy of Forensic Medicine. All rights reserved.

Deep Learning Approaches for Detection of Covid-19 Using Chest X-Ray Images

According to studies, covid-19 affects the respiratory tract along with the lungs. Thus it is critical to efficiently detect and diagnose COVID-19 and the non-COVID i.e., pneumonia/normal cases at the earliest. This helps in reducing the fast spread and stops the condition from becoming virulent. Deep learning models are found to have an extraordinary capacity in providing accurate results, forming an efficient system for detecting COVID-19 as well as pneumonia. Here, 6432 images of chest X-ray, consisting of 5144 images for training and 1288 images for testing, each containing 3 classes - Covid-19 affected, Pneumonia affected and normal were used. The data was preprocessed and a comparison of VGG-16, ResNet-50, and CNN models were done. Models classify data as Covid-19 or Pneumonia or Normal. In the final analysis, the ResNet-50 model gave the highest accuracy of 96.61 percentage followed by VGG-16 with an accuracy of 94.58 percentage, and CNN model with an accuracy of 88.98 percentage. © 2021 IEEE.