Hardhat-Wearing Detection Based on a Lightweight Convolutional Neural Network with Multi-Scale Features and a Top-Down Module

PPE detector: a YOLO-based architecture to detect personal protective equipment (PPE) for construction sites

With numerous countermeasures, the number of deaths in the construction industry is still higher compared to other industries. Personal Protective Equipment (PPE) is constantly being improved to avoid these accidents, although workers intentionally or unintentionally forget to use such safety measures. It is challenging to manually run a safety check as the number of co-workers on a site can be large; however, it is a prime duty of the authority to provide maximum protection to the workers on the working site. From these motivations, we have created a computer vision (CV) based automatic PPE detection system that detects various types of PPE. This study also created a novel dataset named CHVG (four colored hardhats, vest, safety glass) containing eight different classes, including four colored hardhats, vest, safety glass, person body, and person head. The dataset contains 1,699 images and corresponding annotations of these eight classes. For the detection algorithm, this study has used the You Only Look Once (YOLO) family's anchor-free architecture, YOLOX, which yields better performance than the other object detection models within a satisfactory time interval. Moreover, this study found that the YOLOX-m model yields the highest mean average precision (mAP) than the other three versions of the YOLOX.

SHEL5K: An Extended Dataset and Benchmarking for Safety Helmet Detection

Wearing a safety helmet is important in construction and manufacturing industrial activities to avoid unpleasant situations. This safety compliance can be ensured by developing an automatic helmet detection system using various computer vision and deep learning approaches. Developing a deep-learning-based helmet detection model usually requires an enormous amount of training data. However, there are very few public safety helmet datasets available in the literature, in which most of them are not entirely labeled, and the labeled one contains fewer classes. This paper presents the Safety HELmet dataset with 5K images (SHEL5K) dataset, an enhanced version of the SHD dataset. The proposed dataset consists of six completely labeled classes (helmet, head, head with helmet, person with helmet, person without helmet, and face). The proposed dataset was tested on multiple state-of-the-art object detection models, i.e., YOLOv3 (YOLOv3, YOLOv3-tiny, and YOLOv3-SPP), YOLOv4 (YOLOv4 and YOLOv4_pacsp-x-mish), YOLOv5-P5 (YOLOv5s, YOLOv5m, and YOLOv5x), the Faster Region-based Convolutional Neural Network (Faster-RCNN) with the Inception V2 architecture, and YOLOR. The experimental results from the various models on the proposed dataset were compared and showed improvement in the mean Average Precision (mAP). The SHEL5K dataset had an advantage over other safety helmet datasets as it contains fewer images with better labels and more classes, making helmet detection more accurate.

100+ FPS detector of personal protective equipment for worker safety: A deep learning approach for green edge computing

In industrial production, personal protective equipment (PPE) protects workers from accidental injuries. However, wearing PPE is not strictly enforced among workers due to all kinds of reasons. To enhance the monitoring of workers and thus avoid safety accidents, it is essential to design an automatic detection method for PPE. In this paper, we constructed a dataset called FZU-PPE for our study, which contains four types of PPE (helmet, safety vest, mask, and gloves). To reduce the model size and resource consumption, we propose a lightweight object detection method based on deep learning for superfast detection of whether workers are wearing PPE or not. We use two lightweight methods to optimize the network structure of the object detection algorithm to reduce the computational effort and parameters of the detection model by 32% and 25%, respectively, with minimal accuracy loss. We propose a channel pruning algorithm based on the BN layer scaling factor γ to further reduce the size of the detection model. Experiments show that the automatic detection of PPE using our lightweight object detection method takes only 9.5 ms to detect a single video frame and achieves a detection speed of 105 FPS. Our detection model has a minimum size of 1.82 MB and a model size compression rate of 86.7%, which can meet the strict requirements of memory occupation and computational resources for embedded and mobile devices. Our approach is a superfast detection method for green edge computing.

Deep Learning on Construction Sites: A Case Study of Sparse Data Learning Techniques for Rebar Segmentation

Recent advances in deep learning models for image interpretation finally made it possible to automate construction site monitoring processes that rely on remote sensing. However, the major drawback of these models is their dependency on large datasets of training images labeled at pixel level, which must be produced manually by skilled personnel. To reduce the need for training data, this study evaluates weakly and semi-supervised semantic segmentation models for construction site imagery to efficiently automate monitoring tasks. As a case study, we compare fully, weakly and semi-supervised methods for the detection of rebar covers, which are useful for quality control. In the experiments, recent models, i.e., IRNet, DeepLabv3+ and the cross-consistency training model are compared for their ability to segment rebar covers from construction site imagery with minimal manual input. The results show that weakly and semi-supervised models can indeed rival with the performance of fully supervised models with the majority of the target objects being properly found. This study provides construction site stakeholders with detailed information on how to leverage deep learning for efficient construction site monitoring and weigh preprocessing, training, and testing efforts against each other in order to decide between fully, weakly and semi-supervised training.