Requirements for Parametric Design of Physics-Based Synthetic Data Generation for Learning and Inference of Defect Conditions

Shun Hsiang Hsu, Mani Golparvar-Fard

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

The advancement of Artificial Intelligence (AI)-driven defect detection has already demonstrated promises to boost quality assurance and control, as well as condition assessment in the built environment. However, training defect detection models requires hefty amounts of reality capture data, and labeling is considered expensive. In most cases, such data may not cover all situations of defects. Synthetic data, most recently made with Building Information Models (BIM), is turbocharging model development for learning defect features. Nevertheless, few studies focused on characterizing defects to classify their severity, which is crucial to the condition assessment. To that end, this study explores the requirements for generating synthetic data. Parametric physics-based modeling approaches are carefully examined. Using the underlying geometric properties of such data, the condition of each defect can be determined. The feasibility of synthetic defect data is validated with a case study of crack segmentation using the transformer-based model, SegFormer. Examples of how different scenarios can be generated photo-realistically with the use of physics-based rendering for creating varying geometrical characteristics, appearance, and viewpoints of defects are presented. The generated synthetic crack datasets can successfully be used to train the SegFormer model and reach promising predictions on real crack images.

Original languageEnglish (US)
Title of host publicationAdvanced Technologies, Automation, and Computer Applications in Construction
EditorsJennifer S. Shane, Katherine M. Madson, Yunjeong Mo, Cristina Poleacovschi, Roy E. Sturgill
PublisherAmerican Society of Civil Engineers
Pages436-445
Number of pages10
ISBN (Electronic)9780784485262
DOIs
StatePublished - 2024
Externally publishedYes
EventConstruction Research Congress 2024, CRC 2024 - Des Moines, United States
Duration: Mar 20 2024Mar 23 2024

Publication series

NameConstruction Research Congress 2024, CRC 2024
Volume1

Conference

ConferenceConstruction Research Congress 2024, CRC 2024
Country/TerritoryUnited States
CityDes Moines
Period3/20/243/23/24

ASJC Scopus subject areas

  • Civil and Structural Engineering
  • Building and Construction

Fingerprint

Dive into the research topics of 'Requirements for Parametric Design of Physics-Based Synthetic Data Generation for Learning and Inference of Defect Conditions'. Together they form a unique fingerprint.

Cite this