Adrianto OktavianusPO-HAN CHENJACOB JE-CHIAN LIN2024-12-242024-12-242024-12-01https://www.scopus.com/record/display.uri?eid=2-s2.0-85211107528&origin=resultslisthttps://scholars.lib.ntu.edu.tw/handle/123456789/724291Post-earthquake building recovery is always a crucial task after the strike of an earthquake. Conventionally, such building recovery is time-consuming. The availability of building information modeling (BIM) and deep learning enables the possibility of a more efficient building assessment and rehabilitation planning process. In this paper, an intelligent post-earthquake building recovery system that integrates BIM and deep learning is proposed. Deep learning is used for damage classification and recognition, while BIM provides data on all building elements to support the analyses of the building recovery process. The proposed system is expected to assist engineers in building inspection, structural element assessment, and rehabilitation planning. More specifically, the system combines the visual indicators for structural element assessment and semantic segmentation for damage area estimation to achieve damage classification, with a BIM software plug-in for building recovery plan application. Finally, a case study is presented to validate the implementation of the system framework.falseBuilding information modeling (BIM)Building recoveryDeep learningEarthquakeRecovery plan application[SDGs]SDG3[SDGs]SDG11journal article10.1016/j.jobe.2024.1113662-s2.0-85211107528