Gesture Based Computer Control System

The whole world has been witnessing the gigantic enemy in the form of COVID-19 since March 2020. With its super-fast spread, it has devastated a major part of the world and found to be the most dangerous virus of the 21st Century. All countries went into a lockdown to control the spread of the virus, and the economy dropped down to an all- time low index. The major guideline to avoid the spread of diseases like COVID- 19 at work is avoiding contact with people and their belongings. It is not safe to use computing devices because it may result in the spread of the virus by touching them. This paper presents an Artificial Intelligence- based virtual mouse that detects or recognizes hand gestures to control the various functions of a personal computer. The virtual mouse Algorithm uses a webcam or a built-in camera of the system to capture hand gestures, then uses an algorithm to detect the palm boundaries similar to that of the face detection model of the media pipe face mesh algorithm. After tracing the palm boundaries, it uses a regression model and locates the 21 3D hand-knuckle coordinate points inside the recognized hand/palm boundaries. Once the Hand Landmarks are detected, they are used to call windows Application Programming Interface (API) functions to control the functionalities of the system. The proposed algorithm is tested for volume control and cursor control in a laptop with the Windows operating system and a webcam. The proposedsystem took only 1ms to identify the gestures and control the volume and cursor in real-time. © 2023 IEEE.

A Hybrid technolgy using Machine learning and blockchain technology to prevent Covid

Since the year 2020, there has been an outbreak of the respiratory infection that caused a high peak mortality rate, which has led to an increase in the prevalence of Covid. The unanticipated development of the COVID-19 sickness as well as its unchecked global spread show the limitations of the currently available healthcare systems in their ability to respond to emergencies that harm the general population's health. As a result of cutting-edge technology like AI and biological computing (BC) these issues treated promisingly for the covid pandemic. In particular, BC assist in early detection to aid in the fight against pandemics. With the protocols that have been put in place to avoid infections, including the use of masks, social isolation within a radius of 6 meters, routine testing, and two doses of vaccinations. This system comprises the detection of masks, people, and temperatures, as well as the monitoring of information, tracking of in-person contact, and the present state of a person's medical record. Diseases are now able to be traced, and their transmission can be stopped, thanks to advances in technology and the growing prevalence of smartphone use. Because of the reopening of more economic sectors and the continuous widespread distribution of Covid, it is even more important to ensure that you adhere to the provided instructions in order to avoid contracting an infection. © 2023 IEEE.

Face Mask Detection using Single Shot Multibox Detector and Mobile Net

During the pandemic time government took many safety measures to protect the public at common gathering places. People are insisted on wearing a face mask to protect themselves from COVID. Even then many people were roaming without a mask in public places. The proposed technique to detect the face mask is to identify the person's face with mask and person's face without mask and reporting to the safety officers about the persons without mask for further action. The proposed Face mask detection is developed using the ML technique which can be used to classify the people wearing masks and not wearing masks with the input given to the model. The proposed face mask detector is a one-stage detector that focuses on detecting the face mask alone. This work is implemented using the Tensor flow and Computer vision libraries. NumPy is used for image processing. The data set used in MAFA dataset. The model is trained using this data set to get the accurate results. To enable multiple detection here the single shot with multi box detector is used. The base model used for this process is Mobile Net V2. The proposed model is simple and it can be integrated with several other technologies to provide high accuracy percentage of output in the minimum possible time. © 2022 IEEE.

A Review on Recognition of Disguise Face Mask Using Novel Deep Convolutional Neural Network

Object recognition establishes a connection of different objects present in images or videos. Nowadays, this technology is widely used in transportation management systems, intelligence systems, military equipment acquisition, and also in surgical equipment to obtain a surgical guidance, etc. Wearing a facemask has become a mandate in public places to control the spread of coronavirus. This research study has developed a novel facemask detection model based on a single-shot detector (SSD) to collect real-time images. This process has been implemented in three modules: 1) A network of simple error correction features will be introduced based on SSD and partition in order to achieve a better access speed and satisfy the real-time requirements;2) Feature Enhancement Module (FEM) is used to strengthen the in-depth features learned by CNN models to improve the visibility of minor substances;3) A COVID-19-mask will be finally created by considering a large database of face mask images. Test results generate high accuracy while utilizing real-time acquisition and realization of the proposed algorithm. © 2022 IEEE.

Ethereum and Intelligent Systems Technologies for COVID-19

Beginning in 2020, Covid has increased as a result of a burst put on by a respiratory infection with a substantial peaking fatality rate. The unforeseen occurrence and unchecked global spread of the COVID-19 illness highlight the limitations of current healthcare systems in responding to emergencies affecting public wellness. In these conditions, innovative developments like public blockchain and intelligent systems (AI) have emerged as possible treatments for the covid epidemic. In particular, block chain may help with early identification to combat pandemics. With the measures put in place to prevent infection by wearing masks, social seclusion with a 6m radius, routine testing, and two vaccine doses. This system includes mask measurement, people identification, temp sensors, information tracking, in-person interaction locating, and the current state of a user's medical chart. With the development of technology and increased smartphone usage, illnesses may be tracked and their spread controlled. Considering that the expansion of the business sector's rehabilitation and its continued broad distribution of Covid, it is more crucial to adhere to the instructions to avoid contamination. © 2022 IEEE.

Efficacy Determination of Various Base Networks in Single Shot Detector for Automatic Mask Localisation in a Post COVID Setup

The COVID-19 pandemic is one of the rarest events of global crises where a viral pathogen infiltrates every part of the world, leaving every country face an inevitable threat of having to lock down major cities and economic hubs and put firm restrictions on citizens thus slowing down the economy. The risk of removal of lockdowns is the emergence of new waves of a pandemic causing a surge in new cases. These facts necessitate the containment of the virus when the lockdowns end. Wearing masks in crowded places can help restrict the spread of the virus through minuscule droplets in the air. Through the automatic detection, enumeration, and localisation of masks from closed-circuit television footage, it is possible to keep violations of post-COVID regulations in check. In this paper, we leverage the Single-Shot Detection (SSD) framework through different base convolutional neural networks (CNNs) namely VGG16, VGG19, ResNet50, DenseNet121, MobileNetV2, and Xception to compare performance metrics attained by the different variations of the SSD and determine the efficacies for the best base network model for automatic mask detection in a post COVID world. We find that Xception performs best among all the other models in terms of mean average precision.

Optimized face detector-based intelligent face mask detection model in IoT using deep learning approach

The growth of the "Internet of Medical Things (IoMT)” allows for the collection and processing of data in healthcare systems. At the same time, it is challenging to study the requirements of public health prevention. Here, mask-wearing is considered an efficient preventive measure for avoiding virus transfer. Hence, it is necessary to implement an automated mask identification model to prevent public epidemics. The main scope of the proposed method is to design a face mask detection model with IoT using a "Single Shot Multi-box Detector (SSD)” and a hybrid deep learning method. The novelty of the proposed model is that the enhancement made in the face detection and face classification with the developed ASMFO by optimizing the parameters like the threshold in SSD, steps per execution in ResNet, and learning rate in MobileNet, which makes it more efficient and to perform better the conventional models. Here, the parameter optimization is carried out using a hybrid optimization algorithm named Adaptive Sailfish Moth Flame Optimization (ASMFO). Then, the detected face images are given to the hybrid approach named Hybrid ResMobileNet (HResMobileNet)-based classification, where the parameters are tuned using the same ASMFO algorithm for achieving accurate mask detection results. However, the suggested mask identification model with IoT based on three standard datasets is compared with the conventional meta-heuristic algorithms and existing classifiers with various measures. Thus, the experimental analysis is conducted to analyze the effectiveness of the proposed framework over different meta-heuristic algorithms and existing classifiers. The implemented ASMFO-HResMobileNet provides 18.57%, 15.67%, 17.56%, 16.24%, and 19.2% elevated accuracy than SVM, CNN, VGG16-LSTM, ResNet 50, MobileNetv2, and ResNet 50-MobileNetv2. © 2022 Elsevier B.V.

Real-Time Mask Detection Based on SSD-MobileNetV2

After the outbreak of COVID-19, mask detection, as the most convenient and effective means of prevention, plays a crucial role in epidemic prevention and control. An excellent automatic real-time mask detection system can reduce a lot of work pressure for relevant staff. However, by analyzing the existing mask detection approaches, we find that they are mostly resource-intensive and do not achieve a good balance between speed and accuracy. And there is no perfect face mask dataset at present. In this paper, we propose a new architecture for mask detection. Our system uses SSD as the mask locator and classifier, and further replaces VGG-16 with MobileNetV2 to extract the features of the image and reduce a lot of parameters. Therefore, our system can be deployed on embedded devices. Transfer learning methods are used to transfer pre-trained models from other domains to our model. Data enhancement methods in our system such as MixUp effectively prevent overfitting. It also effectively reduces the dependence on large-scale datasets. By doing experiments in practical scenarios, the results demonstrate that our system performed well in real-time mask detection. © 2022 IEEE.

Toward masked face recognition: An effective facial feature extraction and refinement model in multiple scenes

With the impact of the COVID‐19 epidemic, the demand for masked face recognition technology has increased. In the process of masked face recognition, some problems such as less feature information and poor robustness to the environment are obvious. The current masked face recognition model is not quantified enough for feature extraction, there are large errors for faces with high similarity, and the categories cannot be clustered during the detection process, resulting in poor classification of masks, which cannot be well adapted to changes in multiple environments. To solve current problems, this paper designs a new masked face recognition model, taking improved Single Shot Multibox Detector (SSD) model as a face detector, and replaces the input layer VGG16 of SSD with Deep Residual Network (ResNet) to increase the receptive field. In order to better adapt to the network, we adjust the convolution kernel size of ResNet. In addition, we fine‐tune the Xception network by designing a new fully connected layer, and reduce the training cycle. The weights of the three input samples including anchor, positive and negative are shared and clustered together with triplet network to improve recognition accuracy. Meanwhile, this paper adjusts alpha parameter in triplet loss. A higher value of alpha can improve the accuracy of model recognition. We further adopt a small trick to classify and predict face feature vectors using multi‐layer perceptron (MLP), and a total of 60 neural nodes are set in the three neural layers of MLP to get higher classification accuracy. Moreover, three datasets of MFDD, RMFRD and SMFRD are fused to obtain high‐quality images in different scenes, and we also add data augmentation and face alignment methods for processing, effectively reducing the interference of the external environment in the process of model recognition. According to the experimental results, the accuracy of masked face recognition reaches 98.3%, it achieves better results compared with other mainstream models. In addition, the hyper‐parameters tuning experiment is carried out to improve the utilization of computing resources, which shows better results than the indicators of different networks.

A Deep Learning Approach for Detecting Face Mask Using an Improved Yolo-V2 With Squeezenet

Face mask detection has become a critical issue in security and Covid-19 prevention. In this regard, the YOLO V2 network has demonstrated outstanding performance. The YOLO V2 on the other hand, employed Darknet as a feature extractor. However, as compared to Darknet, SqueezeNet allows us to reduce model size while reaching or surpassing the highest accuracy score. SqueezeNet is designed to have lower parameters that can be more readily stored in computer memory and transferred across a computer network. As a result, in this study, we recommended enhancing the YOLO network by replacing Darknet with Squeezenet. Compared to other existing face mask recognition systems that use the standard YOLO V2 algorithm, this improves overall performance in terms of model size and accuracy. As a result, this study proposed a rapid face mask detection model by improving the existing YOLO V2 network architecture by employing logistic classifiers and SqueezeNet for multi-label classification using FMD and MMD face-masked dataset. The model was evaluated on MATLAB 2021 against state-of-the-art approaches. The proposed model outperforms previous algorithms by attaining a good accuracy value of 81% and a recall value of 99.99%. © 2022 IEEE.

Real-Time Mask Detection Based on SSD-MobileNetV2

After the outbreak of COVID-19, mask detection, as the most convenient and effective means of prevention, plays a crucial role in epidemic prevention and control. An excellent automatic real-time mask detection system can reduce a lot of work pressure for relevant staff. However, by analyzing the existing mask detection approaches, we find that they are mostly resource-intensive and do not achieve a good balance between speed and accuracy. And there is no perfect face mask dataset at present. In this paper, we propose a new architecture for mask detection. Our system uses SSD as the mask locator and classifier, and further replaces VGG-16 with MobileNetV2 to extract the features of the image and reduce a lot of parameters. Therefore, our system can be deployed on embedded devices. Transfer learning methods are used to transfer pre-trained models from other domains to our model. Data enhancement methods in our system such as MixUp effectively prevent overfitting. It also effectively reduces the dependence on large-scale datasets. By doing experiments in practical scenarios, the results demonstrate that our system performed well in real-time mask detection. © 2022 IEEE.

Optimized face detector-based intelligent face mask detection model in IoT using deep learning approach

The growth of the "Internet of Medical Things (IoMT)” allows for the collection and processing of data in healthcare systems. At the same time, it is challenging to study the requirements of public health prevention. Here, mask-wearing is considered an efficient preventive measure for avoiding virus transfer. Hence, it is necessary to implement an automated mask identification model to prevent public epidemics. The main scope of the proposed method is to design a face mask detection model with IoT using a "Single Shot Multi-box Detector (SSD)” and a hybrid deep learning method. The novelty of the proposed model is that the enhancement made in the face detection and face classification with the developed ASMFO by optimizing the parameters like the threshold in SSD, steps per execution in ResNet, and learning rate in MobileNet, which makes it more efficient and to perform better the conventional models. Here, the parameter optimization is carried out using a hybrid optimization algorithm named Adaptive Sailfish Moth Flame Optimization (ASMFO). Then, the detected face images are given to the hybrid approach named Hybrid ResMobileNet (HResMobileNet)-based classification, where the parameters are tuned using the same ASMFO algorithm for achieving accurate mask detection results. However, the suggested mask identification model with IoT based on three standard datasets is compared with the conventional meta-heuristic algorithms and existing classifiers with various measures. Thus, the experimental analysis is conducted to analyze the effectiveness of the proposed framework over different meta-heuristic algorithms and existing classifiers. The implemented ASMFO-HResMobileNet provides 18.57%, 15.67%, 17.56%, 16.24%, and 19.2% elevated accuracy than SVM, CNN, VGG16-LSTM, ResNet 50, MobileNetv2, and ResNet 50-MobileNetv2.

A Real-Time Unmasked Detection Using SSD-MobileNetV2 on Edge Device for the COVID-19 Pandemic

COVID-19 Pandemic affects daily life and the global economy. The COVID-19 virus can be spread by small liquid particles, which can be filtered using a face mask. Wearing masks in public areas is an excellent approach to preventing illness. As a result, mask detection is necessary to stop the spread of the disease before a person enters the facility. Regarding Single Shot Multibox Detector-MobileNetV2 (SSD-MobileNetV2) was used in this research to build tools to detect and monitor unmasked people in the facility or working rooms that consist of many people. In this paper, we showed the experimental performance of SSDMobileNetv2 based on an application that runs on an edge device to detect unmasked people in the room and compromise with very high accuracy of 97% in rooms smaller than 16 square meters, which is sufficient to detect the wearing of masks in public places or various locations. © 2022 IEEE.

Face detection for rail transit passengers based on single shot detector and active learning.

COVID-19 spreads rapidly among people, so that more and more people are wearing masks in rail transit stations. However, the current face detection algorithms cannot distinguish between a face wearing a mask and a face not wearing a mask. This paper proposes a face detection algorithm based on single shot detector and active learning in rail transit surveillance, effectively detecting faces and faces wearing masks. Firstly, we propose a real-time face detection algorithm based on single shot detector, which improves the accuracy by optimizing backbone network, feature pyramid network, spatial attention module, and loss function. Subsequently, this paper proposes a semi-supervised active learning method to select valuable samples from video surveillance of rail transit to retrain the face detection algorithm, which improves the generalization of the algorithm in rail transit and reduces the time to label samples. Extensive experimental results demonstrate that the proposed method achieves significant performance over the state-of-the-art algorithms on rail transit dataset. The proposed algorithm has a wide range of applications in rail transit stations, including passenger flow statistics, epidemiological analysis, and reminders of passenger who do not wear masks. Simultaneously, our algorithm does not collect and store face information of passengers, which effectively protects the privacy of passengers.

AI-based Screening System for COVID-19

The outbreak of the COVID-19 pandemic caused by the novel coronavirus has disrupted global health systems and changed the way of life. Leveraging the potential of technology has become indispensable to ensure safety as new strains emerge. In this paper, we propose a low-cost AI-based screening system that can be installed at various locations. The objective of our solution is to automate the task of face mask detection, checking for social distancing, and body temperature scanning. The dataset used for our AI model consisted of 2314 images combining those without a mask and those with an artificial mask attached using computer vision techniques. These three functionalities were carried out by combining AI and IoT technologies. Specifically, we employed the SingleShot-Multibox detector (SSD) and ResNet-10 architecture as the first part of our model for face detection. MobileNetV2 architecture was used as the second part of the model for binary classification (with or without mask). Various IoT components were integrated to achieve a real-time screening process and subsequently simulate entry access or denial. Our proposed model outperformed other existing solutions by achieving an accuracy of 99% and an F1 score of 0.99. © 2022 IEEE.

SSDMNV2-FPN: A Cardiac Disorder Classification from 12 lead ECG images using Deep Neural Network

In this research work, detection of cardio diseases using Object detection techniques from 12 Lead ECG images is proposed. To detect the object in an image with different aspect ratios and with different sizes is one of the main challenges, this issue may lead to the wrong prediction of the diseases. To overcome those challenges, MobileNet with Feature Pyramid Network (FPN) feature extractor is used to extract the feature maps in different aspect ratios. By using the feature maps, the object is detected using the Single Shot Detector (SSD) technique. In addition, weighted sigmoid Focal Loss is adopted to diminish the imbalance among foreground and background samples to enrich detector outcomes. To endorse the effectiveness of the method proposed, a dataset is collected are Abnormal Heart Beat, Covid, Myocardial Infarction, Normal and Previous History of MI. Using the dataset collected, the proposed method gives a mAP accuracy of 95.88% in detection.

Leveraging Discriminative Cues for Masked Face Recognition in Post COVID World

The post COVID world has completely disrupted our lifestyle, where wearing a mask is necessary to protect ourselves and others from contracting the virus. However, face masks have proved to be challenging for facial biometric systems, in the sense that these systems do not work as expected when wearing masks as nearly half of the face is covered, thus reducing discriminative features that the model can leverage. Most of the existing frameworks rely on the entire face as the input, but as the face is covered, these frameworks do not perform up to the mark. Moreover, training another facial recognition system with mask images is challenging as the availability of datasets is limited, both qualitatively and quantitatively. In this paper, we propose a framework that shows better results without significant training. In the proposed work, firstly we extracted the face using SSD, then by obtaining Facial Landmarks for utilizing the cues from other dis-criminative parts for facial recognition. The proposed framework is able to out-perform other frameworks on facial mask images and also found ~4.5% increment in accuracy. © 2022, The Author(s), under exclusive license to Springer Nature Switzerland AG.

Smart Blended Learning Framework based on Artificial Intelligence using MobileNet Single Shot Detector and Centroid Tracking Algorithm

The Covid-19 pandemic has affected all aspects of human life and has even forced humans to shift their life habits, including in the world of education. The learning model must shift from the traditional face-to-face pattern to a modern face-to-face pattern or an asynchronous pattern with information technology-based applications. Blended learning is one of the appropriate solutions to adjust the limited face-to-face learning conditions. Blended learning can be done, for example, by scheduling learning by dividing the number of participants by 50% and entering on a scheduled basis. However, the problem is that the time and effort used are less efficient. Blended learning can also be done by conducting learning simultaneously with 50% of students in class and the remaining 50% through conferences. This concept will streamline the time and effort used. However, the problem is that there is a gap in the learning experience between students in class and students who do learning via conference. This innovative blended learning system framework is proposed to overcome these problems. The system built seeks to present an online learning experience atmosphere so that it is expected to be able to resemble an offline learning atmosphere. We created a system using camera technology and object detection that will track the movement of the teacher so that the teacher can move freely in the room without having to be stuck in front of the computer holding the conference. The algorithms used are MobileNet Single Shot Detector and Centroid Tracking. This research produces an accurate model for detecting teacher movement at a distance of 2, 4, and 6 meters with a camera installation height of 1.5 and 3 meters.

COVID-19: Machine learning for safe transportation.

Entire world has been affected by Covid-19 pandemic. In fighting against the Covid-19, social distancing and face mask have a paramount role in freezing the spread of the disease. People are asked to limit their interactions with each other, to reduce the spread of the disease. Here an alert system has to be maintained to caution people traveling in vehicles. Our proposed solution will work primarily on computer vision. The video stream is captured using a camera. Footage is processed using single shot detector algorithm for face mask detection. Second, YOLOv3 object detection algorithm is used to detect if social distancing is maintained or not inside the vehicle. If passengers do not follow the safety rules such as wearing a mask at any point of the time in the whole journey, alarm/alert is given via buzzer/speaker. This ensures that people abide by the safety rules without affecting their daily norms of transportation. It also helps the government to keep the situation under control.

A Deep Learning Based Novel Approach for Detection of Face Mask Wearing using Enhanced Single Shot Detector (SSD) over Convolutional Neural Network (CNN) with Improved Accuracy

The objective is to build an efficient face mask detector using Single Shot Detector (SSD). The algorithm used for face mask detection was a novel SSD and with the comparison of Convolutional Neural Network (CNN). The face mask detection dataset was usedand the ability of the algorithm was measured with the sample size of 136. SSD has achieved accuracy of 92.25% and for CNN it was 82.6%. By using a base architecture of VGG-16, SSD was able to outperform other object detectors like CNN without compromising speed and accuracy. The SSD and CNN are statistically satisfied with the independent sample t-test value (p<0.05) with a confidence level of 95%. Face mask detection using SSD was significantly better accurate than CNN. © 2022 IEEE.