Determination of workers' compliance to safety regulations using a spatio-temporal graph convolution network

Lee, Bogyeong; Hong, Sungkook; Kim, Hyunsoo

doi:10.1016/j.aei.2023.101942

Cited by 10 publications

(3 citation statements)

References 60 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…[ 63 ]. So OpenPose was widely used in construction sites, where complex behaviors existed and the worker’s body was heavily occluded [ 64 , 65 ]. Therefore, YOLO and OpenPose were selected in this study and were recommended computer vision-based technologies for object identification and motion capture, respectively, at least in the application scenarios similar to this study.…”

Section: Discussionmentioning

confidence: 99%

“…Cao et al [ 21 ] identified miners’ unsafe behavior (10 different types of behaviors) based on ST-GCN in their self-built dataset, with an overall identification accuracy of 86.7%. Lee et al [ 65 ] used ST-GCN to identify 5 different unsafe behaviors of workers, with an overall identification accuracy of 87.20%. The motions in the above studies were quite different in motion characteristics.…”

Section: Discussionmentioning

confidence: 99%

See 1 more Smart Citation

Study on the Interaction Behaviors Identification of Construction Workers Based on ST-GCN and YOLO

Xue

et al. 2023

Sensors

View full text Add to dashboard Cite

The construction industry is accident-prone, and unsafe behaviors of construction workers have been identified as a leading cause of accidents. One important countermeasure to prevent accidents is monitoring and managing those unsafe behaviors. The most popular way of detecting and identifying workers’ unsafe behaviors is the computer vision-based intelligent monitoring system. However, most of the existing research or products focused only on the workers’ behaviors (i.e., motions) recognition, limited studies considered the interaction between man-machine, man-material or man-environments. Those interactions are very important for judging whether the workers’ behaviors are safe or not, from the standpoint of safety management. This study aims to develop a new method of identifying construction workers’ unsafe behaviors, i.e., unsafe interaction between man-machine/material, based on ST-GCN (Spatial Temporal Graph Convolutional Networks) and YOLO (You Only Look Once), which could provide more direct and valuable information for safety management. In this study, two trained YOLO-based models were, respectively, used to detect safety signs in the workplace, and objects that interacted with workers. Then, an ST-GCN model was trained to detect and identify workers’ behaviors. Lastly, a decision algorithm was developed considering interactions between man-machine/material, based on YOLO and ST-GCN results. Results show good performance of the developed method, compared to only using ST-GCN, the accuracy was significantly improved from 51.79% to 85.71%, 61.61% to 99.11%, and 58.04% to 100.00%, respectively, in the identification of the following three kinds of behaviors, throwing (throwing hammer, throwing bottle), operating (turning on switch, putting bottle), and crossing (crossing railing and crossing obstacle). The findings of the study have some practical implications for safety management, especially workers’ behavior monitoring and management.

show abstract

Section: Discussionmentioning

confidence: 99%

Section: Discussionmentioning

confidence: 99%

Study on the Interaction Behaviors Identification of Construction Workers Based on ST-GCN and YOLO

Xue

et al. 2023

Sensors

View full text Add to dashboard Cite

show abstract

“…Estimation of knee angles using computer vison method requires three main processes: pose estimation, landmarks coordinate extraction, relative angle calculation. Pose estimation is performed using OpenPose framework, a deep learningbased system that uses convolutional neural networks to estimate the pose of a person in the video [9], its architecture employed is illustrated in Figure 2. It utilized the Visual Geometry Group (VGG)-19 algorithm within OpenPose.…”

Section: Computer Vision-based Knee Angle Estimationmentioning

confidence: 99%

Computer Vision-Based Approach for Automated Monitoring and Assessment of Gait Rehabilitation at Home

Talaa,

El Fezazi,

Jilbab

et al. 2023

Int. J. Onl. Eng.

View full text Add to dashboard Cite

This study presents a markerless video-based human gait analysis system for automatic assessment of at-home rehabilitation. A marker-based MoCap system (Vicon) is used to evaluate the accuracy of the proposed approach. Additionally, a novel gait rehabilitation score based on the Dynamic Time Warping (DTW) algorithm is introduced, enabling quantification of rehabilitation progress. The accuracy of the proposed approach is assessed by comparing it to a marker-based MoCap system (Vicon), which is used to evaluate the proposed approach. This evaluation results in mean absolute errors (MAE) of 4.8° and 5.2° for the left knee, and 5.9° and 5.7° for the right knee, demonstrating an acceptable accuracy in knee angle measurements. The obtained scores effectively distinguish between normal and abnormal gait patterns. Subjects with normal gait exhibit scores around 97.5%, 98.8%, while those with abnormal gait display scores around 30%, 29%, respectively. Furthermore, a subject at an advanced stage of rehabilitation achieved a score of 65%. These scores provide valuable insights for patients, allowing them to assess their rehabilitation progress and distinguish between different levels of gait recovery. The proposed markerless approach demonstrates acceptable accuracy in measuring knee joint angles during a sagittal walk and provides a reliable rehabilitation score, making it a convenient and cost-effective alternative for automatic at-home rehabilitation monitoring.

show abstract

A Survey of Artificial Intelligence for Industrial Detection

Li,

Hai,

Yin

2024

Ann. Data. Sci.

View full text Add to dashboard Cite

Determination of workers' compliance to safety regulations using a spatio-temporal graph convolution network

Cited by 10 publications

References 60 publications

Study on the Interaction Behaviors Identification of Construction Workers Based on ST-GCN and YOLO

Study on the Interaction Behaviors Identification of Construction Workers Based on ST-GCN and YOLO

Computer Vision-Based Approach for Automated Monitoring and Assessment of Gait Rehabilitation at Home

A Survey of Artificial Intelligence for Industrial Detection

Contact Info

Product

Resources

About