Article Preview
Top1. Introduction
Covid-19 presents an unprecedented and unparalleled challenge to the food systems, public health, and disrupted work by creating unemployment. As the number of cases has increased, the healthcare system has been overwhelmed and has largely failed. WHO has announced this as a pandemic (Denison, 2004). One of the prevention techniques is wearing a mask? It is for your safety as well as for those around you. The virus that causes COVID-19 can be communicated by sneezing, coughing, or even speaking at close range, therefore wearing a mask is necessary. However, if people aren’t following the protocol and not wearing masks then that increases the risk of a community spreading which can become deathly. Many nations have begun the vaccine procedure, but not every country in the globe has access to it. So, until the virus is eradicated, wearing masks daily is a must-do to assist prevent the spread of infection.
Figure 1. A poll was conducted by Morning Consult for people who wear masks.
The figure 1 Poll shows that more than 50% of the population are reluctant to wear masks which is a serious concern and poses a huge threat. Thus, it is very crucial for people to wear masks in public and for the concerned authorities to ensure that people follow the guidelines and the protocols for their safety. Figure 2 below shows the person wearing a mask (A) and the person not wearing a Mask (B).
The primary cause of the infection being intensified was the attention to detail on the part of the population, as well as their lack of awareness. The world is flooded with ordinary individuals, who are either in offices or other residences without wearing any face masks. It is almost impossible to cover everything all the time and on top of that, quite time-consuming. This study is mostly aimed at finding a solution to this problem and helping individuals to protect themselves. This concept is especially relevant in COVID-19, as it is imperative that we preserve ourselves from other people. In addition to offering an effective means to reduce airborne virus infections, the masks will effectively disrupt these infections so that airborne viruses cannot reach a human being's respiratory system, making it an inexpensive way to minimize mortality and respiratory infection disorders. Wearing a facemask can prevent a global pandemic. To avoid this, it is essential to develop an automatic detection for wearing a facemask that will protect each player and prevent the pandemic.
With the recent advancement in deep learning that incorporates computer vision, we are seeing a multitude of breakthroughs in various fields of technology. Deep neural networks' innate capacity to extract information from input photos has recently led to their astounding success in automatic image analysis. Deep learning methods aid in picture classification, quantification, and pattern recognition have identified and emphasized numerous uses of deep neural networks in image processing(Deng et al., 2009; Howard et al., 2017). Convolutional neural networks (CNNs) are the most researched and widely utilized among the several forms of deep neural networks in image processing(Du et al., 2019; Brown et al., 2018; Shin et al., 2016). Although CNN’s have a vast number of parameters, large size annotated datasets are required. Large-scale datasets are often difficult to get by. As a result, researchers are combining CNNs with transfer learning to address this problem. The image representations learned with CNNs on large-scale datasets are efficiently and successfully transferred to other tasks that need small-scale datasets in transfer learning. Given the significant time and computational resources required to develop neural network models from the root of these problems, as well as the significant improvements in a skill that they provide on related problems, using pre-trained models as the starting point in natural language processing and computer vision tasks is an accepted and popular approach in deep learning(Pan & Yang, 2010).