Abstract The spatial characteristics of cracks are significant indicators to assess and evaluate the health existing buildings infrastructures. However, current manual crack description method is time consuming labor consuming. To improve efficiency inspection, advanced computer vision‐based techniques have been utilized detect automatically at image level grid‐cell level. But detections (high specificity) low generality inefficient, in terms that conventional approaches unable identify...

Abstract Recording workers’ activities is an important, but burdensome, management task for site supervisors. The last decade has seen a growing trend toward vision‐based activity recognition. However, recognizing in far‐field surveillance videos understudied. This study proposes hierarchical statistical method videos. consists of two steps. First, deep action recognition was used to recognize actions, and new fusion strategy proposed consider the characteristics with achieved comparable...

It is currently a prevailing trend to adopt clean energies instead of traditional ones due the global climate change caused by carbon emissions. Offshore wind farms, in particular, have emerged as crucial source renewable energy, owing their benefits such no land occupation and abundant resources. However, design, installation, operation, maintenance (O&M) floating turbines (FWTs) involve multiple sources heterogeneous data, which pose challenges data integration management, well simulation...

Construction workers are commonly subjected to ergonomic risks. Accurate assessment is needed reduce However, the diverse and dynamic nature of construction sites makes it difficult collect posture data for without intrusiveness. Therefore, this paper proposed a joint-level vision-based tool (JVEC) provide automatic detailed assessments based on videos. JVEC extracts workers' skeleton from videos with advanced deep learning methods, then Rapid Entire Body Assessment (REBA) used conduct...

Design for Safety (DfS) has been considered as an effective approach to improving construction safety performance via taking into account problems during design. The approach, however, is not yet implemented effectively and efficiently in practice. This research tries develop a practical DfS which can automatically identify potential resulted from design by integrating BIM (Building Information Modeling) with rules. rules related defined built based on regulations or codes, integrated...

Workers' unsafe behaviors are one of the main causes for construction accidents. Fully understanding on site will help to prevent them, thus reducing The accurate and timely identification workers' can aid in analysis prevention However, traditional methods (e.g., observation) behavior data collection is neither efficient nor comprehensive. This paper develops a skeleton-based real-time method by combining image-based technologies, safety knowledge, ergonomic theory. proposed recognizes...

Factorization Machines (FMs) refer to a class of general predictors working with real valued feature vectors, which are well-known for their ability estimate model parameters under significant sparsity and have found successful applications in many areas such as the click-through rate (CTR) prediction. However, standard FMs only produce single fixed representation each across different input instances, may limit CTR model’s expressive predictive power. Inspired by success Input-aware (IFMs),...

Overexertion-related construction activities are identified as a leading cause of work-related musculoskeletal disorders (WMSDs) among workers. However, few studies have focused on the automated recognition overexertion-related workers’ well assessing ergonomic risk levels, which may help to minimize WMSDs. Therefore, this study examined feasibility using acceleration and foot plantar pressure distribution data captured by wearable insole system for levels. The proposed approach was tested...

Several studies have analyzed heart rate variability (HRV) using nonlinear methods, such as approximate entropy, the largest Lyapunov exponent, and correlation dimension in patients with cardiovascular disorders. However, few used methods to analyze HRV order determine level of physical fatigue experienced by construction workers. As a result, identify categorize workers, current study examined linear approaches analysis. Fifteen healthy workers (mean age, 33.2±6.9 years) were selected for...

Construction workers are commonly subject to ergonomic risks due awkward working postures or lifting/carrying heavy objects. Accordingly, accurate assessment is needed help improve efficiency and reduce risks. However, the diverse dynamic nature of construction activities makes it difficult unobtrusively collect worker behavior data for analysis. To address this issue, an automatic workload approach proposed first time continuously assess body joints using image-based three-dimensional (3D)...

Factorization machines (FMs) are a class of general predictors working effectively with sparse data, which represents features using factorized parameters and weights. However, the accuracy FMs can be adversely affected by fixed representation trained for each feature, as same feature is usually not equally predictive useful in different instances. In fact, inaccurate may even introduce noise degrade overall performance. this work, we improve explicitly considering impact individual input...

Abstract Automatic activity recognition plays an important role in addressing the efficiency issue of site management. In recent years, there has been increasing interest vision‐based recognition, while its relatively low accuracy and speed impede practical application. This paper introduces a discriminative model to combine deep features contextual information improve activities workers on foot surveillance videos. Specifically, conditional random field (CRF) is designed based features,...