Masked Face Recognition Model with Explainable AI

Due to the recent COVID-19 pandemic, people tend to wear masks indoors and outdoors. Therefore, systems with face recognition, such as FaceID, showed a tendency of decline in accuracy. Consequently, many studies and research were held to improve the accuracy of the recognition system between masked faces. Most of them targeted to enhance dataset and restrained the models to get reasonable accuracies. However, not much research was held to explain the reasons for the enhancement of the accuracy. Therefore, we focused on finding an explainable reason for the improvement of the model's accuracy. First, we could see that the accuracy has actually increased after training with a masked dataset by 12.86%. Then we applied Explainable AI (XAI) to see whether the model has really focused on the regions of interest. Our approach showed through the generated heatmaps that difference in the data of the training models make difference in range of focus. © 2023, The Author(s), under exclusive license to Springer Nature Switzerland AG.

Efficient intelligent compression and recognition system-based vision computing for computed tomography COVID images

Computed tomography (CT) image-based medical recognition is extensively used for COVID recognition as it improves recognition and scanning rate. A method for intelligent compression and recognition system-based vision computing for CT COVID (ICRS-VC-COVID) was developed. The proposed system first preprocesses lung CT COVID images. Segmentation is then used to split the image into two regions: nonregion of interest (NROI) with fractal lossy compression and region of interest with context tree weighting lossless. Subsequently, a fast discrete curvelet transform (FDCT) is applied. Finally, vector quantization is implemented through the encoder, channel, and decoder. Two experiments were conducted to test the proposed ICRS-VC-COVID. The first evaluated the segmentation compression, FDCT, wavelet transform, and discrete curvelet transform (DCT). The second evaluated the FDCT, wavelet transform, and DCT with segmentation. It demonstrates a significant improvement in performance parameters, such as mean square error, peak signal-to-noise ratio, and compression ratio. At similar computational complexity, the proposed ICRS-VC-COVID is superior to some existing techniques. Moreover, at the same bit rate, it significantly improves the quality of the image. Thus, the proposed method can enable lung CT COVID images to be applied for disease recognition with low computational power and space. © The Authors. Published by SPIE under a Creative Commons Attribution 4.0 International License. Distribution or reproduction of this work in whole or in part requires full attribution of the original publication, including its DOI. [DOI: 10.1117/1.JEI.32.2.021404] © 2023 SPIE. All rights reserved.

Non-Contact Physiological Measurement System for Wearing Masks During the Epidemic

Physiological signals indicate a person's physical and mental state at any given time. Accordingly, many studies extract physiological signals from the human body with non-contact methods, and most of them require facial feature points. However, under COVID-19, wearing a mask has become a must in many places, so how non-contact physiological information measurements can still be performed correctly even when a mask covers the facial information has become a focus of research. In this study, RGB and thermal infrared cameras were used to execute non-contact physiological information measurement systems for heart rate, blood pressure, respiratory rate, and forehead temperature for people wearing masks due to the pandemic. Using the green (G) minus red (R) signal in the RGB image, the region of interest (ROI) is established in the forehead and nose bridge regions. The photoplethysmography (PPG) waveforms of the two regions are obtained after the acquired PPG signal is subjected to the optical flow method, baseline drift calibration, normalization, and bandpass filtering. The relevant parameters in Deep Neural Networks (DNN) for the regression model can correctly predict the heartbeat and blood pressure. In addition, the temperature change in the ROI of the mask after thermal image processing and filtering can be used to correctly determine the number of breaths. Meanwhile, the thermal image can be used to read the temperature average of the ROI of the forehead, and the forehead temperature can be obtained smoothly. The experimental results show that the above-mentioned physiological signals of a subject can be obtained in 6-s images with the error for both heart rate and blood pressure within 2%∼3% and the error of forehead temperature within ±0.5°C. © 2023 Tech Science Press. All rights reserved.

Region-of-Interest Aware 3D ResNet for Classification of COVID-19 Chest Computerised Tomography Scans

Coronavirus disease 2019, commonly known as COVID-19, is an extremely contagious disease caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2). Computerised Tomography (CT) scans based diagnosis and progression analysis of COVID-19 have recently received academic interest. Most algorithms include two-stage analysis where a slice-level analysis is followed by the patient-level analysis. However, such an analysis requires labels for individual slices in the training data. In this paper, we propose a single-stage 3D approach that does not require slice-wise labels. Our proposed method comprises volumetric data pre-processing and 3D ResNet transfer learning. The pre-processing includes pulmonary segmentation to identify the regions of interest, volume resampling and a novel approach for extracting salient slices. This is followed by proposing a region-of-interest aware 3D ResNet for feature learning. The backbone networks utilised in this study include 3D ResNet-18, 3D ResNet-50 and 3D ResNet-101. Our proposed method employing 3D ResNet-101 has outperformed the existing methods by yielding an overall accuracy of 90%. The sensitivity for correctly predicting COVID-19, Community Acquired Pneumonia (CAP) and Normal class labels in the dataset is 88.2%, 96.4% and 96.1%, respectively. © 2013 IEEE.

Face mask detection using deep convolutional neural network and multi-stage image processing

• A customized image dataset is built for research on face mask detection. • The dataset is manually labeled to provide high annotation accuracy. • For Face mask detection customized CNN with multi-step image processing is used. • The performance of the proposed CNN is compared with YOLO v3 and Faster R-CNN. • Two publicly available datasets including MAFA and MOXA used for validation. Face mask detection has several applications including real-time surveillance, biometrics, etc. Face mask detection is also useful for surveillance of the public to ensure face mask wearing in public places. Ensuring that people are wearing a face mask is not possible with monitoring staff;instead, automatic systems are a much better choice for face mask detection and monitoring to help manage public behaviour and contribute to restricting the outbreak of COVID-19. Despite the availability of several such systems, the lack of a real image dataset is a big hurdle to validating state-of-the-art face mask detection systems. In addition, using the simulated datasets lack the analysis needed for real-world scenarios. This study builds a new dataset namely RILFD by taking real pictures using a camera and annotating them with two labels (with mask, without mask) which are publicly available for future research. In addition, this study investigates various machine learning models and off-the-shelf deep learning models YOLOv3 and Faster R-CNN for the detection of face masks. The customized CNN models in combination with the 4 steps of image processing are proposed for face mask detection. The proposed approach outperforms other models and proved its robustness with a 97.5% of accuracy score in face mask detection on the RILFD dataset and two publicly available datasets (MAFA and MOXA). [ FROM AUTHOR] Copyright of Image & Vision Computing is the property of Elsevier B.V. and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full . (Copyright applies to all s.)

Deep Crowd Analysis to Spot Social Distancing Violations in Post-COVID 19 Lifestyle

Crowd analysis is a new field of study that involves processing a large group of people to examine one or more of their behaviors. Deep learning is an appropriate technique for crowd analysis using a convolutional neural network. To calculate the distance between crowd members and to identify social distance violations, a deep crowd analysis is proposed in this study. Pre-trained in a single class To discover the region of interest, CNN is utilised to classify people (RoI). The people in the picture are then localized using a density map. The reference point used to calculate the distance between the people is the centroid of the isolated areas in the density map. A social distance violation is reported if the estimated distance is less than the specified threshold distance (3 meters). Between the two ROIs, a distance measured in pixels is determined. © 2022 IEEE.

Validation of a Deep Learning Model for Detecting Chest Pathologies from Digital Chest Radiographs.

Purpose: Manual interpretation of chest radiographs is a challenging task and is prone to errors. An automated system capable of categorizing chest radiographs based on the pathologies identified could aid in the timely and efficient diagnosis of chest pathologies. Method: For this retrospective study, 4476 chest radiographs were collected between January and April 2021 from two tertiary care hospitals. Three expert radiologists established the ground truth, and all radiographs were analyzed using a deep-learning AI model to detect suspicious ROIs in the lungs, pleura, and cardiac regions. Three test readers (different from the radiologists who established the ground truth) independently reviewed all radiographs in two sessions (unaided and AI-aided mode) with a washout period of one month. Results: The model demonstrated an aggregate AUROC of 91.2% and a sensitivity of 88.4% in detecting suspicious ROIs in the lungs, pleura, and cardiac regions. These results outperform unaided human readers, who achieved an aggregate AUROC of 84.2% and sensitivity of 74.5% for the same task. When using AI, the aided readers obtained an aggregate AUROC of 87.9% and a sensitivity of 85.1%. The average time taken by the test readers to read a chest radiograph decreased by 21% (p < 0.01) when using AI. Conclusion: The model outperformed all three human readers and demonstrated high AUROC and sensitivity across two independent datasets. When compared to unaided interpretations, AI-aided interpretations were associated with significant improvements in reader performance and chest radiograph interpretation time.

Cost-Effective Solution of Remote Photoplethysmography Capable of Real-Time, Multi-Subject Monitoring with Social Distancing.

Recent advances in remote-photoplethysmography (rPPG) have enabled the measurement of heart rate (HR), oxygen saturation (SpO2), and blood pressure (BP) in a fully contactless manner. These techniques are increasingly applied clinically given a desire to minimize exposure to individuals with infectious symptoms. However, accurate rPPG estimation often leads to heavy loading in computation that either limits its real-time capacity or results in a costly setup. Additionally, acquiring rPPG while maintaining protective distance would require high resolution cameras to ensure adequate pixels coverage for the region of interest, increasing computational burden. Here, we propose a cost-effective platform capable of the real-time, continuous, multi-subject monitoring while maintaining social distancing. The platform is composed of a centralized computing unit and multiple low-cost wireless cameras. We demonstrate that the central computing unit is able to simultaneously handle continuous rPPG monitoring of five subjects with social distancing without compromising the frame rate and rPPG accuracy.

Explainable Diagnosis, Lesion Segmentation and Quantification of COVID-19 Infection from CT Images Using Convolutional Neural Networks

The health crisis caused by the COVID-19 pandemic has led to unprecedented research efforts to build AI solutions that can assist healthcare systems. In this work, we propose a novel CNN-based system that detects COVID-19 infection and other pneumonia from CT scans, segments COVID-specific lesions, namely Ground Glass Opacities (GGO) and Consolidations (CL), and computes the percentage of lungs that have been affected by COVID and provides an explanation of the basis in which the diagnosis has been made through the comparison of class activation maps pertaining to the diagnosis with the segmented lesions. This can assist healthcare setups in the rapid contactless screening of COVID-19 and assess the stage and severity of the disease, while also providing some level of transparency on the rationale behind the model's decisions. Based on the initial results of the interpretation of the model's decisions, all the non-lung areas from the CT images were removed using a contour detection-based region of interest (ROI) extraction approach. This was done to prevent the model from making decisions based on details in non-lung areas, which are clinically irrelevant for COVID diagnosis. This is the first work to utilize such a contour detection-based ROI extraction approach for medical images, based on our study. The model has achieved a mean F1 score of 0.87 for multi-label classification (COVID, Common Pneumonia & Normal) and a Dice Similarity Coefficient (DSC) of 0.8066 for lesion segmentation which has exceeded the DSC achieved by 6 out of 7 lesion segmentation models referenced in our study. © 2022 IEEE.

Image Encryption Based on 5-Neighbor Programmable Cellular Automata

With the rapid development and spread of non-face-to-face digital technologies and services due to the spread of COVID-19, real-time non-face-to-face online meetings, education, telemedicine, online collaboration, telecommuting, various non-face-to-face tasks in the financial sector, and the sharing economy are frequent. As a result, network traffic increases and the demand for real-time security of multimedia is increasing. Security of information, including image content, is an essential part of today's communication technology and is very important for safe transmission. In this paper, we design a 5-neighbor programmable cellular automata (FNPCA) based ROI (region of interest) image encryption system that can effectively reduce computational cost and maintain an appropriate level of security. © 2022 IEEE.

Prospective evaluation of thoracic diseases using a compact flat-panel detector spiral computed tomographic scanner.

Objective: To prospectively evaluate the image quality and diagnostic performance of a compact flat-panel detector (FD) scanner for thoracic diseases compared to a clinical CT scanner. Materials and methods: The institutional review board approved this single-center prospective study, and all participants provided informed consent. From December 2020 to May 2021, 30 patients (mean age, 67.1 ± 8.3 years) underwent two same-day low-dose chest CT scans using clinical state-of-art and compact FDCT scanners. Image quality was assessed visually and quantitatively. Two readers evaluated the diagnostic performance for nodules, parenchymal opacifications, bronchiectasis, linear opacities, and pleural abnormalities in 40 paired CT scans. The other 20 paired CT scans were used to examine the agreement of semi-quantitative CT scoring regarding bronchiectasis, bronchiolitis, nodules, airspace consolidations, and cavities. Results: FDCT images had significantly lower visual image quality than clinical CT images (all p < 0.001). The two CT image sets showed no significant differences in signal-to-noise and contrast-to-noise ratios (56.8 ± 12.5 vs. 57.3 ± 15.2; p = 0.985 and 62.9 ± 11.7 vs. 60.7 ± 16.9; p = 0.615). The pooled sensitivity was comparable for nodules, parenchymal opacifications, linear opacities, and pleural abnormalities (p = 0.065-0.625), whereas the sensitivity was significantly lower in FDCT images than in clinical CT images for micronodules (p = 0.007) and bronchiectasis (p = 0.004). The specificity was mostly 1.0. Semi-quantitative CT scores were similar between the CT image sets (p > 0.05), and intraclass correlation coefficients were around 0.950 or higher, except for bronchiectasis (0.869). Conclusion: Compact FDCT images provided lower image quality but comparable diagnostic performance to clinical CT images for nodules, parenchymal opacifications, linear opacities, and pleural abnormalities.

Remote Pulse Estimation in the Presence of Face Masks

Remote photoplethysmography (rPPG), a family of techniques for monitoring blood volume changes, may be especially useful for contactless health monitoring via face videos from consumer-grade cameras. The COVID-19 pandemic caused widespread use of protective face masks, which results in a domain shift from the typical region of interest. In this paper we show that augmenting unmasked face videos by adding patterned synthetic face masks forces the deep learning-based rPPG model to attend to the periocular and forehead regions, improving performance and closing the gap between masked and unmasked pulse estimation. This paper offers several novel contributions: (a) deep learning-based method designed for remote photoplethysmography in a presence of face masks, (b) new dataset acquired from 54 masked subjects with recordings of their face and ground-truth pulse waveforms, (c) data augmentation method to add a synthetic mask to a face video, and (d) evaluations of handcrafted algorithms and two 3D convolutional neural network-based architectures trained on videos of unmasked faces and with masks synthetically added. © 2022 IEEE.

Detection of Face Mask Using Convolutional Neural Network

According to data acquired by the World Health Organization, the worldwide universal of COVID-19 bears harshly hit the realm and bears immediately contaminate eight heaps of human beings in general. Wearing face masks and following cautious public leave behind are two of the embellished protection from harm rules of conduct that need to take the place of honestly held places in consideration of keeping from happening or continuing the spread of the virus. To develop in mind or physically conservative surroundings that contribute to public protection from harm, we suggest an adept data processing machine located in close contact with the genuine in existence-period made or done by a human being to discover two reliable public dissociate themselves and face masks honestly placed by the model ahead of the start of the model to monitor special interests or pursuits and discover rape through photographic equipment. In addition to presenting an alarm to the public, in this proposed structure, we have designed mask detection along which indicates people to wear their mask properly before permitting in to the area which they prefer. We have used machine learning with supports the accuracy for the prediction. © 2022 IEEE.

A Region-based Compression Technique for Medical Image Compression using Principal Component Analysis (PCA)

Region-based compression technique is particularly useful for radiological archiving system as it allows diagnostically important regions to be compressed with near lossless quality while the non-diagnostically important regions (NROI) to be compressed at lossy quality. In this paper, we present a region-based compression technique tailored for MRI brain scans. In the proposed technique termed as automated arbitrary PCA (AAPCA), an automatic segmentation based on brain symmetrical property is used to separate the ROI from the background. The arbitrary-shape ROI is then compressed by block-to-row PCA algorithm (BTRPCA) based on a factorization approach. The ROI is optimally compressed with lower compression rate while the NROI is compressed with higher compression rate. The proposed technique achieves satisfactory segmentation performance. The subjective and objective evaluation performed confirmed that the proposed technique achieves better performance metrics (PSNR and CoC) and higher overall compression rate. The experimental results also demonstrated that the proposed technique is more superior to various state-of-the-art compression methods.

Deep Learning-Based Face Mask Detecting System: An Initiative Against COVID-19

In this era of pandemic, wearing mask and taking precaution is a must, and this paper provides a way of detecting mask with the help of deep learning. Deep learning is an important part of machine learning. OpenCV, an important part of Python, is used for real-time image detection. This is a very decent system which can be applied at various platforms and can help in slowing down the transmission of coronavirus. © 2022, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.

Social Distancing Crowd Segmentation, Estimation and Visualisation

This research shows a modern crowd counting solution which alters typical prediction solutions into a segmentation of individuals based on a distance threshold, allowing for better visualisation and results. The study proposes using YOLOv4-normal and YOLOv4-tiny models, which have shown great results throughout calibration with an MAE of 14 and 36 respectively. However it did present some issues of accuracy degradation when trained on head annotations at any level of crowd density. As for visualisation, perspective transformation was used which directly helped in providing the distance calculation that was absent from standard transformation. If any variants of YOLOv4 are to be used, the main argument is the choice between speed over accuracy while relying on native implementations. In the case of distance regulation, any transformation that maps itself onto the region of interest, such as perspective transformation should be used to precisely determine distances from a camera to the region of interest itself. © 2021 IEEE.

A Grad-CAM and YOLO based approach to extracting a human passageway ROI

We propose a computer vision based solution for the use of public video feeds to monitor crowd congestion with a focus on full automation as a potential scalable solution to address crowd statistics extraction needs amplified by the COVID-19 pandemic. The novelty is the provision of a fully autonomous solution that is able to generate a region of interest (ROI) upon initial feed registration with a self-refinement algorithm that perfects the ROI over time. Five classes were used from the Places 2 dataset. The root model of the hierarchy was used to classify between a beach, fast-food restaurant, train station, lawn and market with an overall accuracy of 95.58% and F1-Score of 88.94%. The market and beach class were then split into two sub-classes each. The 'beach' model was further explored using a Grad-CAM based post-processing technique to better understand what the model bases the classification on. The novelty is the use of the same technique to generate a human passageway region of interest based on the localisation of the Grad-CAM in several live beach footages. These were also inferred using a YOLOv5 based human tracking approach. The Grad-CAM based ROI was then evaluated for each footage on the YOLOv5 generated ROI. © 2021 IEEE.

Contactless Palm Vein Biometric Authentication and Health Monitoring Device

Today, the novel COVID-19 pandemic has taken a significant toll on countries worldwide and impacted the lives and well-being of people across nations. Measures need to be taken to slow down the spread of such viruses, which can be reduced by taking proper precautions to avoid unnecessary contact and incurring hygienic habits. People can become infected by touching infected objects or surfaces, then touching their eyes, nose or mouth. We need proper and hygienic authentication systems for granting access to authorized users at various places such as companies, universities, banks, etc. Today there are multiple biometric systems that can serve this purpose, but to maintain hygiene we need systems that are contactless in nature thus to reduce spread of infections through touch. The palm vein pattern is distinctive biometric identity of individuals that is also a safe and reliable biometric authentication technique. The major advantage of palm vein biometric that we consider in our proposed system is that this kind of authentication can be done in a contactless way. Here, we propose an authentication system that uses palm vein pattern biometric to authenticate users in a contactless way. We have also added the functionalities of temperature detection and blood oxygen level detection to this system. These health symptoms are the major symptoms of various fatal virus infections such as the coronavirus. By checking the users for these symptoms before granting access, we can further limit the spread of infections and also help detect the infected patients. The proposed system does this in a contactless way. © 2021 IEEE.

Particle Swarm Optimization for Adaptive Social-distance of Neighborhood in the IoT and COVID-19 Era

While it is well understood that the emerging Social Internet of Things (SIoT) offers a description of a new world of billions of humans which are intelligently communicate and interact with each other. SIoT presents new challenges for suggesting useful objects with certain services for people. This is due to the limitation of social networks between human and objects, such as the evaluation of the various patterns inherent in human walk in cities. In this study we focus services on the problem of recommendation on SIoT which is very important for many applications such as urban computing, smart cities, and health care. The optimized results of swarm of certain infected people COViD-19 introduced in this paper aims at finding a given region of interest. Guided by a fitness function, the particle swarm optimization (PSO) algorithm has proved its efficiency to explore the search space and find the optimal solution. However, in real world scenarios in which the peoples are simulated as particles, there are practical constraints that should be taken into considerations. The most two significant constraints are (1) given the social-distance, the measurement of input variable fluctuations and their possibility of occurring via probability distribution function over the whole particles. (2) given the limited the communication range of particle/people/users, therefore, the spread of the diseases are simulated and evaluated using neighborhood particle swarm optimization (NPSO). © 2021 IEEE.

Detection of Social Distance and Intimation System for Covid-19

COVID-19’s is a novel corona virus, fast spread has caused substantial damage and affected more than tens of millions of individuals around the world. People frequently wear masks to safeguard themselves and others against the transmission of coronavirus. The world health organization conveys the people to follow the social distancing to prevent the spread of Covid. Researchers have proposed several machine learning models to classify the disease, but none have identified the algorithm which gives more accuracy. Also, similar studies that have proposed various other techniques for prediction. In addition to that maintaining social distancing is also a major factor. In these regions, personally monitoring whether individuals are maintaining social distancing or not is quite impossible. This study aims to develop a highly accurate and real-time technique for the automatic detection of individuals who are not maintaining social distancing. Three state-of-the-art object identification models, namely YOLOv4, Tiny-YOLOv4 are used to detect the objects. Many results suggest that YOLO v4 has the greatest mAP value of 88.90%, followed by YOLO v4 and Tiny-YOLO v4 with mAP values of 82.24% and 74.80%, respectively. © 2022, The Author(s), under exclusive license to Springer Nature Switzerland AG.