Generalization of Construction Object Segmentation Models using Self-Supervised Learning

Yeji Hong, Wei Chih Chern, Tam Nguyen, Hongjo Kim

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

Abstract

To evaluate the safety of construction site workers, deep learning models recognizing workers and safety equipment in construction site images are widely used. However, it is frequently observed that deep learning models based on supervised learning methods do not work well for unseen data in other domains having different visual characteristics. To address this issue, a novel method for generalizing semantic segmentation models was proposed. This method adopts two strategies: a domain adaptation method based on self-supervised learning and a copy-paste data augmentation. Source domain data with annotations (workers and hardhats) and target domain data without annotations are used for model training in a self-supervised learning scheme. The proposed model showed an improved generalization capability in semantic segmentation without annotation data of the target domain.

Original languageEnglish
Title of host publicationProceedings of the 38th International Symposium on Automation and Robotics in Construction, ISARC 2021
EditorsChen Feng, Thomas Linner, Ioannis Brilakis
PublisherInternational Association for Automation and Robotics in Construction (IAARC)
Pages843-848
Number of pages6
ISBN (Electronic)9789526952413
Publication statusPublished - 2021
Event38th International Symposium on Automation and Robotics in Construction, ISARC 2021 - Dubai, United Arab Emirates
Duration: 2021 Nov 22021 Nov 4

Publication series

NameProceedings of the International Symposium on Automation and Robotics in Construction
Volume2021-November
ISSN (Electronic)2413-5844

Conference

Conference38th International Symposium on Automation and Robotics in Construction, ISARC 2021
Country/TerritoryUnited Arab Emirates
CityDubai
Period21/11/221/11/4

Bibliographical note

Funding Information:
Project for Smart Construction Technology (No.21SMIP-A158708-02)” funded by the Korea Agency for Infrastructure Technology Advancement under the Ministry of Land, Infrastructure and Transport, and managed by the Korea Expressway Corporation.

Funding Information:
This research was conducted with the support of the “2021 Yonsei University Future-Leading Research Initiative (No.2021-22-0037)” and the “National R&D

Publisher Copyright:
© 2021 Proceedings of the International Symposium on Automation and Robotics in Construction. All rights reserved.

All Science Journal Classification (ASJC) codes

  • Building and Construction
  • Civil and Structural Engineering
  • Control and Systems Engineering
  • Safety, Risk, Reliability and Quality
  • Artificial Intelligence
  • Computer Science Applications

Fingerprint

Dive into the research topics of 'Generalization of Construction Object Segmentation Models using Self-Supervised Learning'. Together they form a unique fingerprint.

Cite this