The present study proposes a Real-Time Pose Estimation technique using OpenPose based on ResNet-50 that enables rapid safety prevention and accident detection among field workers. Field workers perform tasks in high-risk environments, and accurate Pose Estimation is a crucial aspect of ensuring worker safety. However, it is difficult for Real-Time Pose Estimation to be conducted in such a way as to simultaneously meet Real-Time processing requirements and accuracy in complex environments. To address these issues, the current study uses the OpenPose algorithm based on ResNet-50, which is a neural network architecture that performs well in both image classification and feature extraction tasks, thus providing high accuracy and efficiency. OpenPose is an algorithm specialized for multi-human Pose Estimation that can be used to estimate the body structure and joint positions of a large number of individuals in real time. Here, we train ResNet-50-based OpenPose for Real-Time Pose Estimation and evaluate it on various datasets, including actions performed by real field workers. The experimental results show that the proposed algorithm achieves high accuracy in the Real-Time Pose Estimation of field workers. It also provides stable results while maintaining a fast image processing speed, thus confirming its applicability in real field environments.
The advent of the Fourth Industrial Revolution has revolutionized the manufacturing sector by integrating artificial intelligence into vision inspection systems to improve the efficiency and quality of products. Supervised-learning-based vision inspection systems have emerged as a powerful tool for automated quality control in various industries. During visual inspection or final inspection, a human operator physically inspects a product to determine its condition and categorize it based on their know-how. However, the know-how-based visual inspection process is limited in time and space and is affected by many factors. High accuracy in vision inspection is highly dependent on the quality and precision of the labeling process. Therefore, supervised learning methods of 1-STAGE DETECTION, such as You Only Look Once (YOLO), are utilized in automated inspection to improve accuracy. In this paper, we proposed a labeling method that achieves the highest inspection accuracy among labeling methods such as NG intensity and NG intensity when performing anomaly detection using YOLOv7 in the cosmetics manufacturing process.
In this study, we present the few-shot generative adversarial network (FS-GAN) model, which integrates few-shot learning and a generative adversarial network with an unsupervised learning approach (AnoGAN) to address the challenges of anomaly detection in smart-factory manufacturing environments. Manufacturing processes often encounter malfunctions or defective parts that disrupt production and compromise product quality. However, collecting and labeling sufficient data to detect anomalies is time-intensive, and abnormal data are rare, leading to data imbalances. The FS-GAN model leverages few-shot learning to enable accurate predictions with minimal data and uses the generative capabilities of AnoGAN to mitigate the scarcity of abnormal data by generating synthetic normal data. Experimental results demonstrate that FS-GAN outperforms existing models in terms of accuracy and learning speed, even with limited datasets, effectively addressing the data imbalance problem inherent in manufacturing. The model reduces dependency on extensive data collection and labeling efforts, making it suitable for real-world applications. Through reliable and efficient anomaly detection, FS-GAN contributes to production reliability, product quality, and operational efficiency in smart factories. This study highlights the potential of FS-GAN to provide a cost-effective and high-performance solution to the challenges of anomaly detection in the manufacturing industry.
Industrial sites pose numerous hazards where unexpected accidents can occur at any time, and personal protective equipment (PPE) is a primary safeguard for worker safety. In this study, PPE specifically refers to safety helmets, safety shoes, and safety gloves. Manual verification of PPE usage is infeasible in environments with many workers, motivating automated detection. This study proposes a method that integrates the Convolutional Block Attention Module (CBAM) exclusively into the training-only auxiliary reversible branch of YOLOv9’s Programmable Gradient Information (PGI) architecture. The proposed CBAMLinear module enhances gradient information during training while introducing zero additional computational overhead at inference, as the entire auxiliary branch is removed. The proposed YOLOv9 with CBAMLinear achieved consistent mAP@0.5:0.95 gains of 0.005–0.007 over the baseline for the three larger model variants, while maintaining identical inference-time parameters and FLOPs. In industrial safety, even modest performance gains can directly contribute to accident prevention by reducing false positives and false negatives, making this approach well suited for real-time safety management systems in industrial settings.
Detecting anomalies in time-series data has significant implications in various domains, including the early detection of equipment and facility conditions in industrial settings and the identification of risk factors in financial markets. However, sensor-based time-series data obtained from real-world sites predominantly consist of normal patterns, with anomalies being extremely rare, resulting in insufficient data for model training. In this paper, we present a method that addresses the imbalance between normal and anomaly data by transforming time-series data into a structured sentence format and using a large language model to generate and augment diverse anomaly scenarios. The generated sentences are then converted back into time-series data and subjected to an internal validation process. This approach enables the construction of a broader range of anomaly data than existing techniques.The proposed method is designed for univariate, continuous valued time-series and is anticipated to serve as an effective data augmentation tool, with extension to broader domains left to future work.