A Naive and Semantic Approach for Detecting Face Mask Region Based Convolutional Neural Networks (R-CNN)

In this paper, we investigate the problem of face mask compliance classification in response to the Corona Virus Disease (COVID-19) pandemic. Since the start of the pandemic, many governments and businesses have been continually updating policies to help slow the spread of the virus, including requiring face masks to use many public and private services. In response to these policies, many researchers have developed new face detection and recognition techniques for masked faces. Many of the developed approaches proposed to address the problem of masked face detection have been relatively successful. However, these approaches almost exclusively focus on detecting the presence or absence of someone wearing a face mask. It is understood and broadly discussed in various reports and the media that there are people not always following the suggested guidelines provided by public health authorities for wearing face masks, which include ensuring the face mask properly covers the nose and mouth. To date, very few research publications exist that investigate the capability of modern classification algorithms to efficiently distinguish between masked faces in the visible band that are either compliant or non-compliant with the suggested guidelines. Furthermore, to the best of our knowledge, there is no publication in the open literature that focuses on the investigation of face mask classification in the thermal band. As thermal sensors continue to improve in quality and decrease in cost, surveillance applications using thermal imagery are expected to continue to grow and benefit those organizations considering the automation of face mask detection and compliance. In this study, we propose an investigation on face mask compliance in both the thermal and visible bands. It is composed of the following salient steps; (1) the creation of a multi-spectral masked face database from subjects wearing or not wearing face masks, (2) the augmentation of the generated database with synthetic face masks to simulate two different levels of non-compliant wearing of face masks, and (3) the assessment of a variety of CNN architectures on the previous augmented database to investigate any differences between classifying thermal and visible masked faces. Experimental results show that face mask compliance classification in both studied bands yield a classification accuracy that reaches 100% for most models studied, when experimenting on frontal face images captured at short distances with adequate illumination.KeywordsFace masksCOVID-19PandemicFace mask classificationDeep learningMask complianceThermal imagingMulti-spectral Imaging

The face mask is the first line of defense against infectious particulates and droplets that may cause illness. Currently in the Philippines, the wearing of face mask is compulsory whenever citizens leave their residences as mandated by the government to mitigate the spread of COVID-19. The wearing of face masks has become a new normal among Filipinos. This created market opportunities for different types which became commonly and immediately available for purchase. This study aimed to differentiate the effectiveness of locally available face masks in terms of electrostatic filtration capability. Twelve different types of face masks grouped into five categories – surgical, fabric, N95 variants, foam type, and novelty type – were evaluated. Electrostatic fields were measured from each face mask including pore sizes via scanning electron microscopy. Moreover, by utilizing the estimated charge and mass of the SARS-CoV-2 virion, the transmission rate was simulated using COMSOL Multiphysics®. It was observed that face masks with negatively charged materials combined with small pore sizes afforded less particle transmission. The results of this study are of timely significance in potentially laying out public awareness in the selection and utilization of face masks that can provide foremost shielding against viral transmission.

Uncovering the release of micro/nanoplastics from disposable face masks at times of COVID-19

Computer vision learning pay a high attention due to global pandemic COVID-19 to enhance public health service. During the fatality, tiny object detection is a more challenging task of computer vision, as it recruits the pair of classification and detection beneath of video illustration. Compared to other object detection deep neural networks demonstrated a helpful object detection with a superior achievement that is Face mask detection. However, accession with YOLOv3 covered by an exclusive topic which through certainly happening natural disease people get advantage. Added with face mask detection performed well by the YOLOv3 where it measures real time performance regarding a powerful GPU. whereas computation power with low memory YOLO darknet command sufficient for real time manner. Regarding the paper section below we have attained that people who wear face masks or not, its trained by the face mask image and non face mask image. Under the experimental conditions, real time video data that finalized over detection, localization and recognition. Experimental results that show average loss is 0.0730 after training 4000 epochs. After training 4000 epochs mAP score is 0.96. This unique approach of face mask visualization system attained noticeable output which has 96% classification and detection accuracy.

Face masks have been associated with effective prevention of diffusion of viruses via droplets. However, the use of face masks among children, especially those aged younger than 3 years, is debated, and the US Centers for Disease Control and American Academy of Physicians recommend the use of face mask only among individuals aged 3 years or older. To examine whether the use of surgical facial masks among children is associated with episodes of oxygen desaturation or respiratory distress. This cohort study was conducted from May through June 2020 in a secondary-level hospital pediatric unit in Italy. Included participants were 47 healthy children divided by age (ie, group A, aged ≤24 months, and group B, aged >24 months to ≤144 months). Data were analyzed from May through June 2020. All participants were monitored every 15 minutes for changes in respiratory parameters for the first 30 minutes while not wearing a surgical face mask and for the next 30 minutes while wearing a face mask. Children aged 24 months and older then participated in a walking test for 12 minutes. Changes in respiratory parameters during the use of surgical masks were evaluated. Among 47 children, 22 children (46.8%) were aged 24 months or younger (ie, group A), with 11 boys (50.0%) and median (interquartile range [IQR]) age 12.5 (10.0-17.5) months, and 25 children (53.2%) were aged older than 24 months to 144 months or younger, with 13 boys (52.0%) and median (IQR) age 100.0 (72.0-120.0) months. During the first 60 minutes of evaluation in the 2 groups, there was no significant change in group A in median (IQR) partial pressure of end-tidal carbon dioxide (Petco2; 33.0 [32.0-34.0] mm Hg; P for Kruskal Wallis = .59), oxygen saturation (Sao2; 98.0% [97.0%-99.0%]; P for Kruskal Wallis = .61), pulse rate (PR; 130.0 [115.0-140.0] pulsations/min; P for Kruskal Wallis = .99), or respiratory rate (RR; 30.0 [28.0-33.0] breaths/min; P for Kruskal Wallis = .69) or for group B in median (IQR) Petco2 (36.0 [34.0-38.0] mm Hg; P for Kruskal Wallis = .97), Sao2 (98.0% [97.0%-98.0%]; P for Kruskal Wallis = .52), PR (96.0 [84.0-104.5] pulsations/min; P for Kruskal Wallis test = .48), or RR (22.0 [20.0-25.0] breaths/min; P for Kruskal Wallis = .55). After the group B walking test, compared with before the walking test, there was a significant increase in median (IQR) PR (96.0 [84.0-104.5] pulsations/min vs 105.0 [100.0-115.0] pulsations/min; P < .02) and RR (22.0 [20.0-25.0] breaths/min vs 26.0 [24.0-29.0] breaths/min; P < .05). This cohort study among infants and young children in Italy found that the use of facial masks was not associated with significant changes in Sao2 or Petco2, including among children aged 24 months and younger.

Objectives: Prevention of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) and the disease COVID-19 is a public health priority. The efficacy of non-pharmaceutical interventions such as wearing face masks to prevent SARS-CoV-2 infection has been well established in controlled settings. However, evidence for the effectiveness of face masks in preventing SARS-CoV-2 transmission within real-world settings is limited and mixed. The present systematic review evaluated the effectiveness of face mask policies and mask wearing to prevent SARS-CoV-2 transmission and COVID-19 in real-world settings. Methods: Following PRISMA guidelines, scientific databases, and gray literature, were searched through June 2023. Inclusion criteria were as follows: (1) studies/reports written in or translated to English; (2) prospectively assessed incidence of SARS-CoV-2 or COVID-19; (3) assessed the behavior and/or policy of mask-wearing; and (4) conducted in community/public settings (i.e., not laboratory). Studies were excluded if they did not parse out data specific to the effect of mask wearing (behavior and/or policy) and subsequent SARS-CoV-2 transmission or COVID-19 disease or if they relied solely on statistical models to estimate the effects of mask wearing on transmission. A total of 2616 studies were initially identified, and 470 met inclusion and exclusion criteria for full-text review. The vote counting method was used to evaluate effectiveness, and risk of bias was assessed using JBI critical appraisal tools. Results: A total of 79 unique studies met the final inclusion criteria, and their data were abstracted and evaluated. Study settings included community/neighborhood settings (n = 34, 43%), healthcare settings (n = 30, 38%), and school/universities (n = 15, 19%). A majority of studies (n = 61, 77%) provided evidence to support the effectiveness of wearing face masks and/or face mask policies to reduce the transmission of SARS-CoV-2 and/or prevention of COVID-19. Effectiveness of mask wearing did not vary substantially by study design (67–100%), type of mask (77–100%), or setting (80–91%), while 85% of masking policies specifically reported a benefit. Conclusions: This systematic literature review supports public health recommendations and policies that encourage the public to wear face masks to reduce the risk of SARS-CoV-2 infection and COVID-19 in multiple real-world settings. Effective communication strategies are needed to encourage and support the use of face masks by the general public, particularly during peak infection cycles.

Community face masking is possibly of great value in reducing COVID-19 transmission, especially when universally adopted with high compliance. The aim of this study is to investigate the knowledge, common misconceptions, barriers, and the compliance of the community with the use of face masks for the prevention of COVID-19. A validated questionnaire was administered to the participants through a web link by using various social media. The collected data were statistically analyzed for significant differences according to demographic variables. The average knowledge of face masks and their role in preventing COVID-19 transmission was 95.64%, with no differences among most of the demographical factors. Older groups and females demonstrated a better attitude towards wearing face masks than other groups did (p<0.001). Another significant difference in the participant’s attitude was noticed between the various educational levels, employment, and nationality (p<0.001). Of the total respondents, 88.2% encouraged wearing face masks. Misconceptions about wearing face masks were very low. The frequency of wearing face masks at public places, workplaces, or social gatherings was 87.2%, 80.5%, and 47.5% respectively. There was a significant variation in the compliance with wearing face masks between the various groups based on age, gender, nationality, and employment status (p<0.001). The inconvenience in wearing face masks was reported by 36.3%. Face irritation and ear pain were reported by 70.2% and 43.5%, respectively. The inconvenience of wearing face masks with eyeglasses was reported by 44.3% of those wearing eyeglasses. In general, the study demonstrated a good attitude among participants towards wearing face masks. Although the respondents in the study were aware of the benefits of wearing face masks, the barriers may have decreased their desire to do so. These barriers include difficulty in breathing, discomfort, face irritation, and ear pain.

The current instance segmentation method can achieve satisfactory results in common scenarios. However, under the overlap or partial occlusion between targets caused by the complex scenes, accurate segmentation of pigs remains a challenging task. To address the problem, the authors propose an instance segmentation method based on Mask Scoring region‐based convolutional neural networks (R‐CNN) (MS R‐CNN), which creates the adversarial network called MaskDis in the head branch of MS R‐CNN. The MaskDis is trained as a discriminator using a generative adversarial network, and the MS R‐CNN model is used as a generator during model training. The adversarial training enables the generator to learn context information and features at the pixel level, which effectively improves the segmentation quality under pigs’ overlapping or dense occlusions scenes. Experimental conducted on the pig object segmentation dataset show that the proposed approach achieves a precision of 92.03%, a recall of 92.18%, and an F1 score of 0.9210. Compared with the basic MS R‐CNN model, the approach achieved a 2.25% improvement in precision and 1.18% improvement in F1 score. Furthermore, the improved approach outperformed advanced instance segmentation methods such as YOLACT, Swin Transformer, YOLOv5‐seg, and SOLOv2 on COCO evaluation metrics. These experimental results demonstrate the effectiveness of the proposed approach in instance segmentation of pigs in complex scenes, providing technical support for non‐contact pig automatic management.

Is there sufficient evidence to suggest that the wearing of face masks is imperative during surgical procedures for maximum patient protection? Or is it just another ritual that needs to be re-examined in the light of today's emphasis on evidence-based medicine. Should we be re-educating theatre staff that the wearing face masks is more for their own protection than for that of the patient? To help us towards some conclusions, Peggy Edwards presents the results of her review of the literature on the face mask debate.

Corona Virus Disease (COVID-19) pandemic is causing a health crisis. One of the effective methods against the virus is wearing a face mask. This paper introduces face mask detection that can be used by the authorities to make mitigation, evaluation, prevention, and action planning against COVID-19. The face mask recognition in this study is developed with a machine learning algorithm through the image classification method: MobileNetV2. The steps for building the model are collecting the data, pre-processing, split the data, testing the model, and implement the model. The built model can detect people who are wearing a face mask and not wearing it at an accuracy of 96,85 percent. After the model implemented in 25 cities from various source of image, the percentage of people wearing face mask in the cities has a strong correlation to the vigilance index of COVID-19 which is 0,62.

Recently the world is suffering from a pandemic. Nations were extensively affected by the Covid-19. Individuals are fighting the battle to survive; people lost their lives or the ones that matter to them while some lost their livelihood. In an hour of need everyone is trying their best to win this battle, (WHO) World Health Organization advised, to maintain at least one-meter distance, to get a vaccination, and to wear face masks. Face masks can be efficacious as a barrier between humans and the virus. The Face Mask Detection performs a critical task in suppressing the transmission of covid-19. Here, the Face masks detector was constructed with a hybrid technique. The real-time face masks detector was distributed into two segments: Training Phase and Testing Phase. The paper represents the union of the transfer learning concept, MobileNetV2 model by google, and the Caffe model, with the help of some open-source libraries used for computer vision and machine learning. The datasets were provided by the Kaggle, and they were parted into two categories: training and testing sets. The model's accuracy was compared for the two datasets. The dataset with a greater number of images has better accuracy, and to avoid overfitting data augmentation was used.

As a consequence of the COVID-19 pandemic, billions of people wear face masks for hours on a daily basis. However, whether hazardous materials released from the face masks might pose potential risks to human health remains mostly unexplored. In particular, nanoparticle-containing face masks have become popular in the market due to their improved antiviral and antibacterial effects. The leachability of nanoparticles and any subsequent toxicity are largely unknown. This study is the first to evaluate the Cu leachability from a face mask containing Cu nanoparticles and the subsequent cytotoxicity of the leachate. We showed that ∼20 μg of Cu in the mask was inhaled, accounting for almost 3% of the total Cu. Moreover, 30–40% of the Cu attached to fibers from the mask could be easily released into solution, and most of the released Cu existed in ionic forms. Cell-based testing indicated that although the leachate did not induce any cytotoxicity, it caused significant Cu stress leading to the exocytosis of lysosomes, changes in glutathione concentration, and the generation of reactive oxygen species. Overall, our results suggest that the leachability of nanoparticle-containing face masks should be carefully assessed, especially considering that most nanoparticle-containing face masks are recommended to be reusable after being washed.

Do Slit-Lamp Shields and Face Masks Protect Ophthalmologists amidst COVID-19?

Accurate estimation of methane (CH4) and carbon dioxide (CO2) emissions requires precise sampling methods due to the significant variability of these gases. This study, part of the project breed4green in Austria, investigated the variability of CH4 and CO2 and minimum number of spot samples required to accurately estimate CH4 and CO2 emissions from dairy cows on commercial farms. A total of 51,895 spot samples from 451 cows on 9 farms were analyzed, which were obtained using the GreenFeed system. Each farm participated in 2 GreenFeed measurement periods (Period 1 and 2), each lasting 6 wk and separated by 3 to 6 mo. A total of 298 cows, each with at least 70 spot samples in either Period 1 or Period 2, were analyzed to determine the minimum number of spot samples. Two methods were used. In the first, forward sampling, subsets of increasing spot samples (the first 1, 10, 20, 30, 40, 50, 60 and 70 measurements per cow) were compared with the complete set of available measurements for the cows. In the second method, random sampling, the same number of spot samples were selected randomly from all available visits for each cow and this procedure was repeated 100 times to account for sampling variability. The minimum number of spot samples required was determined when the Pearson correlation between the arithmetic mean of the subset and that of all available measurements reached a value of 0.90. Mean CH4 and CO2 emissions were 430 g/d and 13,309 g/d, respectively. The coefficient of variation (CV) for diurnal variability ranged from 4.6 to 12.0% for CH4 and from 2.7 to 6.9% for CO2, indicating low to moderate diurnal variability. Day-to-day variability was low, with CV between 3.7 and 8.7% for CH4 and between 2.5 and 8.7% for CO2. Substantial within-cow variation was observed for CH4 (mean CV of 22.5%), while CO2 was more stable (mean CV of 11.7%). When using forward sampling, at least 19 spot samples were required for accurate CH4 estimates and only 13 for CO2, with a mean of 6.5 and 4.3 d needed to obtain these measurements, respectively. When using random sampling, only 8 spot samples for CH4 and 5 spot samples for CO2 were sufficient to meet the accuracy criteria. For future analyses of this data set, we will rely on the results of forward sampling. Forward sampling is based on the actual sequence of visits made by each cow to the GreenFeed and therefore offers a more conservative and realistic approach compared with random sampling. Overall, the results highlight the dynamic nature of CH4 emissions, which are influenced by factors such as feeding patterns and rumen fluctuations and suggest that more spot samples are needed for reliable estimates of CH4 than for CO2.

In order to ensure the safety of transmission lines, the use of unmanned aerial vehicle (UAV) images for automatic object detection has important application prospects, such as the detection of birds’ nests. The traditional bird’s nest detection methods mainly include the study of morphological characteristics of the bird’s nest. These methods have poor applicability and low accuracy. In this work, we propose a deep learning-based birds’ nests automatic detection framework—region of interest (ROI) mining faster region-based convolutional neural networks (RCNN). First, the prior dimensions of anchors are obtained by using k-means clustering to improve the accuracy of coordinate boxes generation. Second, in order to balance the number of foreground and background samples in the training process, the focal loss function is introduced in the region proposal network (RPN) classification stage. Finally, the ROI mining module is added to solve the class imbalance problem in the classification stage, combined with the characteristics of difficult-to-classify bird’s nest samples in the UAV images. After parameter optimization and experimental verification, the deep learning-based bird’s nest automatic detection framework proposed in this work achieves high detection accuracy. In addition, the mean average precision (mAP) and formula 1 (F1) score of the proposed method are higher than the original faster RCNN and cascade RCNN. Our comparative analysis verifies the effectiveness of the proposed method.