RESUMEN
The spread of Corona Virus Disease 19 (COVID-19) in Indonesia is still relatively high and has not shown a significant decrease. One of the main reasons is due to the lack of supervision on the implementation of health protocols such as wearing masks in daily activities. Recently, state-of-the-art algorithms were introduced to automate face mask detection. To be more specific, the researchers developed various kinds of architectures for the detection of masks based on computer vision methods. This paper aims to evaluate well-known architectures, namely the ResNet50, VGG11, InceptionV3, EfficientNetB4, and YOLO (You Only Look Once) to recommend the best approach in this specific field. By using the MaskedFace-Net dataset, the experimental results showed that the EfficientNetB4 architecture has better accuracy at 95.77% compared to the YOLOv4 architecture of 93.40%, InceptionV3 of 87.30%, YOLOv3 of 86.35%, ResNet50 of 84.41%, VGG11 of 84.38%, and YOLOv2 of 78.75%, respectively. It should be noted that particularly for YOLO, the model was trained using a collection of MaskedFace-Net images that had been pre-processed and labelled for the task. The model was initially able to train faster with pre-trained weights from the COCO dataset thanks to transfer learning, resulting in a robust set of features expected for face mask detection and classification.
RESUMEN
In recent years, the performance of people-counting models has been dramatically increased that they can be implemented in practical cases. However, the current models can only count all of the people captured in the inputted closed circuit television (CCTV) footage. Oftentimes, we only want to count people in a specific Region-of-Interest (RoI) in the footage. Unfortunately, simple approaches such as covering the area outside of the RoI are not applicable without degrading the performance of the models. Therefore, we developed a novel learning strategy that enables a deep-learning-based people counting model to count people only in a certain RoI. In the proposed method, the people counting model has two heads that are attached on top of a crowd counting backbone network. These two heads respectively learn to count people inside the RoI and negate the people count outside the RoI. We named this proposed method Gap Regularizer and tested it on ResNet-50, ResNet-101, CSRNet, and SFCN. The experiment results showed that Gap Regularizer can reduce the mean absolute error (MAE), root mean square error (RMSE), and grid average mean error (GAME) of ResNet-50, which is the smallest CNN model, with the highest reduction of 45.2%, 41.25%, and 46.43%, respectively. On shallow models such as the CSRNet, the regularizer can also drastically increase the SSIM by up to 248.65% in addition to reducing the MAE, RMSE, and GAME. The Gap Regularizer can also improve the performance of SFCN which is a deep CNN model with back-end features by up to 17.22% and 10.54% compared to its standard version. Moreover, the impacts of the Gap Regularizer on these two models are also generally statistically significant (P-value < 0.05) on the MOT17-09, MOT20-02, and RHC datasets. However, it has a limitation in which it is unable to make significant impacts on deep models without back-end features such as the ResNet-101.