DenseNet201 Model for Robust Detection on Incorrect Use of Mask

In 2022, the COVID-19 pandemic is still occurring. One of the optimal prevention efforts is to wear a mask properly. Several previous studies have classified the use of masks incorrectly. However, the accuracy resulting from the classification process is not optimal. This research aims to use the transfer learning method to achieve optimal accuracy. In this research, we used three classes, namely without a mask, incorrect mask, and with a mask. The use of these three classes is expected to be more detailed in detecting violations of the use of masks on the face. The classification method used in this research uses transfer learning as feature extraction and Global Average Pooling and Dense layers as classification layers. The transfer learning models used in this research are MobileNetV2, InceptionV3, and DenseNet201. We evaluate the three models' accuracy and processing time when using video data. The experimental results show that the DenseNet201 model achieves an accuracy of 93%, but the processing time per video frame is 0.291 s. In contrast to the MobileNetV2 model, which produces an accuracy of 89% and the processing speed of each video frame is 0.106 s. This result is inversely proportional to accuracy and speed. The DenseNet201 model produces high accuracy but slow processing time, while the MobileNetV2 model is less accurate but has faster processing. This research can be applied in the crowd center to monitor health protocols in the use of masks in the hope of inhibiting the transmission of the COVID-19 virus. © 2023, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.

Covid -19 Social Distance Analysis Using Machine Learning

According to the Ministry of Global Health, social distance is one of the most effective defenses against COVID-19 and helps to prevent its spread. Governments have imposed many safety orders on citizens and facilities to limit social distancing and slow the spread of the virus. As a result, there has been an increase in interest in technologies to research and control the spread of COVID-19 in various settings. This research aims to investigate the results of several machine learning approaches to find cases when the physical distance between people has been violated. The method first identifies the instance of the human in the video frame, tracks the movements, computes the distance with other humans on the same frame and thus estimates the number of people who violate the social distance. Compares the approach to performing the performance using Yolo, SSD and Faster R- CNN. Videos that are used in this approach are collected from the wild, considering different camera settings, indoor and outdoor scenes, and recorded from various angles. Comparing the three methods Yolo, SSD and Faster RNN, the results show Yolo has a better performance in detecting humans from the current videos and thus in determining the violation of the distance between humans. © 2023 IEEE.

Social Distance Alerting System Using Machine Learning

The increase into the Corona virus pandemic led to a higher death rate globally. The best way to prevent getting sick is to keep yourself physically or socially far. Our project provides an approach for physical isolation revealing using machine knowledge toward indicate the necessary space to be maintained to decrease the collision of the corona virus contagious widespread spread. By analyzing a videotape provide for from the camera, the detect apparatus be fashioned in the direction of notify individuals toward maintain a out of harm's way aloofness on or after one an additional. The open-source person recognition pretrained model, YOLO3 algorithm, was utilized to recognize people using the video frame from the camera as input. YOLO3 has the benefit of mortal a lot quicker than further algorithms, at a halt maintain exactness and meets the real-time requirements for person detection. In order to calculate distance from the 2D plane, the video frames are afterwards transformed into top-down views. Estimated distance between individuals and any non-compliant pair of individuals within the display is indicate by means of a red colour edge and stripe, the moderate distance is represented with orange colour and the safe distance is represented by green colour frame. The suggested technique was examined lying on a pre record videotape as well as on the live video feed of persons walking on the road. Additionally an alarm sound is provided to notify the persons. The outcome show that the planned strategy is ready toward sees the societal separation trial among many populaces withinthe videotape. © 2022 IEEE.

Real-Time Face Mask Detection from CCTV Video Frames using Deep Neural Networks

The coronavirus, commonly known as SARS COVID-19, is causing a pandemic that is affecting individuals all over the world. The spread of the virus compelled the authorities to impose a rigorous lockdown on its citizens. Every person in society may experience a variety of issues as a result of this. According to WHO (World Health Organization) regulations, the sole method to halt the virus's spread is to wear a face mask. Therefore, the suggested approach makes sure that everyone appropriately wears a face mask in public locations. The objective of this approach is to detect people without face masks and people who wear facemasks incorrectly in social environments. This system consists of multiple face detection modules to find the area of interest within the video frames. In the next level, using the trained Deep Learning model, the presence of a mask is detected and faces without mask and faces wearing masks incorrectly are highlighted. The dataset for face mask identification comprises of 8190 photos with unique facial annotations from the Kaggle and RMFD datasets that come into two categories: "with mask” and "without mask”. © 2022 IEEE