Relationship Reasoning of Construction Activities Based on Spatiotemporal Attention Mechanisms

DOI

10.2991/978-94-6463-728-1_78How to use a DOI?

Keywords

Smart Construction; Computer Vision; Construction Management; Construction Activity; Relationship Reasoning; Transformer

Abstract

The perception of construction activities is crucial for intelligent construction and management. Visual data acquired through surveillance cameras on construction sites is rich in information but faces challenges such as low information density, complex analysis, and high storage demands. Construction sites are complex environments with cluttered and dynamic visual backgrounds, involving various objects and interactions that complicate intelligent construction understanding. This study deployed a video collection system at a public building construction site in Shanghai to capture standard floor construction activities. By using object detection and re-identification techniques, workers and key construction elements were tracked. The study utilized spatiotemporal attention mechanisms for video feature extraction and combined objects trajectories with category information to reason the relationships between workers and on-site objects, forming subject-relationship-object triplets that describe construction processes. The established visual relationship reasoning models and dataset have proven effective, with accuracy exceeding 90% on the test set. Although lighting and occlusion issues on-site can affect results and the current study does not fully cover the entire construction process, the acquired data can support intelligent construction process simulation, efficiency analysis, construction plan optimization, and progress prediction, demonstrating significant application potential.

Copyright

© 2025 The Author(s)

Open Access

Open Access This chapter is licensed under the terms of the Creative Commons Attribution-NonCommercial 4.0 International License (http://creativecommons.org/licenses/by-nc/4.0/), which permits any noncommercial use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.

Download article (PDF)

TY  - CONF
AU  - Binghan Zhang
AU  - Bin Yang
PY  - 2025
DA  - 2025/05/19
TI  - Relationship Reasoning of Construction Activities Based on Spatiotemporal Attention Mechanisms
BT  - Proceedings of the 3rd International Conference on Green Building, Civil Engineering and Smart City (GBCESC 2024)
PB  - Atlantis Press
SP  - 849
EP  - 855
SN  - 2352-5401
UR  - https://doi.org/10.2991/978-94-6463-728-1_78
DO  - 10.2991/978-94-6463-728-1_78
ID  - Zhang2025
ER  -