Eine Plattform für die Wissenschaft: Bauingenieurwesen, Architektur und Urbanistik
Eine Plattform für die Wissenschaft: Bauingenieurwesen, Architektur und Urbanistik
Informativeness‐guided active learning for deep learning–based façade defects detection
Annotation work is burdensome and challenging for developing a façade defects detector, especially when the raw data set is large but not all useful. To alleviate the problem, this study proposes an informativeness‐guided active learning methodology to effectively select informative data to train a robust façade defects detector. A novel data annotation workflow is developed to ensure the high quality of labels. Then, an active learning–based model training strategy is adopted to enable the model to have both the abilities of generalization and discrimination on different defect features. Besides, an innovative informativeness assessment method is proposed by flexibly combining the degree of uncertainty and the degree of representativeness. Through the proposed method, the performance of façade defects detection can be further boosted with the same amount but more informative training data so that the cost‐efficiency of human annotation work can be improved.
Informativeness‐guided active learning for deep learning–based façade defects detection
Annotation work is burdensome and challenging for developing a façade defects detector, especially when the raw data set is large but not all useful. To alleviate the problem, this study proposes an informativeness‐guided active learning methodology to effectively select informative data to train a robust façade defects detector. A novel data annotation workflow is developed to ensure the high quality of labels. Then, an active learning–based model training strategy is adopted to enable the model to have both the abilities of generalization and discrimination on different defect features. Besides, an innovative informativeness assessment method is proposed by flexibly combining the degree of uncertainty and the degree of representativeness. Through the proposed method, the performance of façade defects detection can be further boosted with the same amount but more informative training data so that the cost‐efficiency of human annotation work can be improved.
Informativeness‐guided active learning for deep learning–based façade defects detection
Guo, Jingjing (Autor:in) / Wang, Qian (Autor:in) / Su, Shu (Autor:in) / Li, Yiting (Autor:in)
Computer‐Aided Civil and Infrastructure Engineering ; 38 ; 2408-2425
01.11.2023
18 pages
Aufsatz (Zeitschrift)
Elektronische Ressource
Englisch
Deep Learning-based Segmentation of Thermal Images for Heat Loss Detection in Building Façade
| Taylor & Francis Verlag | 2024
Deep learning for estimating energy savings of early-stage facade design decisions
| DOAJ | 2021