Change detection network for construction housekeeping using feature fusion and large vision models

Automation in Construction, Journal Year: 2025, Volume and Issue: 172, P. 106038 - 106038

Published: Feb. 6, 2025

Language: Английский

---

SSD-Based Innovations for Improved Construction Management

Journal of Engineering Research, Journal Year: 2025, Volume and Issue: unknown

Published: Feb. 1, 2025

Language: Английский

---

Towards 4D BIM: A Systematic Literature Review on Challenges, Strategies and Tools in Leveraging AI with BIM

Buildings, Journal Year: 2025, Volume and Issue: 15(7), P. 1072 - 1072

Published: March 26, 2025

In moving toward the fourth dimension of building information modeling (4D BIM), this study systematically reviews literature on challenges, strategies, and tools in 4D BIM-related research. To address limitation static nature knowledge represented traditional (BIM), BIM incorporates time into systems to anticipate potential delays, optimize workflows, improve overall project efficiency architecture, engineering, construction (AEC) industry. Although existing research has covered various aspects, in-depth review studies specifically remain scarce. Following a systematic search data analysis, work examines contexts (building models, lean systems, ontology frameworks, predictive tools, software techniques) evaluates them qualitatively. The evaluation identified several key strategies for advancing BIM, including integration methodologies, frameworks. These approaches contribute automation sharing optimization processes within AEC digital infrastructures. This highlights gaps current emphasizes importance integrated solutions while also classifying software, standards related presenting foundation future AI-driven solutions.

Language: Английский

---

UAV-based automated earthwork progress monitoring using deep learning with image inpainting

Automation in Construction, Journal Year: 2025, Volume and Issue: 175, P. 106211 - 106211

Published: April 28, 2025

Language: Английский

---

From raw to refined: Data preprocessing for construction machine learning (ML), deep learning (DL), and reinforcement learning (RL) models

Automation in Construction, Journal Year: 2024, Volume and Issue: 168, P. 105844 - 105844

Published: Oct. 24, 2024

Language: Английский

---

Automatic Vision-Based Dump Truck Productivity Measurement Based on Deep-Learning Illumination Enhancement for Low-Visibility Harsh Construction Environment

Journal of Construction Engineering and Management, Journal Year: 2024, Volume and Issue: 150(11)

Published: Aug. 24, 2024

Language: Английский

---

Spatial Patterns and Multi-Dimensional Impact Analysis of Urban Street Quality Perception under Multi-Source Data: A Case Study of Wuchang District in Wuhan, China

Applied Sciences, Journal Year: 2023, Volume and Issue: 13(21), P. 11740 - 11740

Published: Oct. 26, 2023

The human spatial perception of urban streets has a high complexity and traditional research methods often focus on access surveys perception. Urban serve as both direct conduit for pedestrians’ impressions city reflection the quality that city. Street-view images can provide large amount primary data image semantic segmentation technique. Deep learning techniques were used in this study to collect boring, beautiful, depressing, lively, safe, wealthy scores street spaces based these images. Then, pattern street-space was analyzed by global Moran’s I GIS hotspot analyses. findings demonstrate various facilities affect different ways strength an influencing factor’s influence varies depending its geographical location. results factors reveal difference degree positive negative perceptions visual dimension pedestrians. contribution is it reduces potential bias single source using multi-dimensional impact analysis explain relationship between elements. study’s offer direction high-quality development well advice planning enhanced design.

Language: Английский

---

Object Detection and Instance Segmentation in Construction Sites

Published: March 22, 2024

Identifying safety hazards and ensuring quality control in construction sites requires information from surveillance images, such as clear detection boxes, accurate edges of targets, etc. Although both object instance segmentation models have yielded good results on large open datasets, there is still room for improvement the field. Besides, previous methods based traditional DNNs encountered bottlenecks improving performance, are far enough to be used practice. We present Cascade Mask R-CNN framework, which simultaneously handles tasks scenarios using a single model. And Swin Transformer model serves backbone model, enhances feature maps extracted by strengthens sites. Additionally, we propose post-processing method that refines initial coarse masks generated Our experiments show our highly effective at detecting objects segmenting instances sites, with mean average precision (mAP) 60.2% 54% segmentation.

Language: Английский

---

Synthesizing High-Quality Construction Segmentation Datasets Through Pre-trained Diffusion Model

Lecture notes in computer science, Journal Year: 2024, Volume and Issue: unknown, P. 354 - 366

Published: Jan. 1, 2024

Language: Английский

---

Advancements in 3D digital model generation for digital twins in industrial environments: Knowledge gaps and future directions

Advanced Engineering Informatics, Journal Year: 2024, Volume and Issue: 62, P. 102929 - 102929

Published: Oct. 1, 2024

Language: Английский

---