“…In recent years, both DL‐based and non‐DL‐based computer vision methods have been proposed for damage detection of different types of structural systems, such as reinforced concrete structures (Gao & Mosalam, 2018; Liang, 2019; Pan & Yang, 2020; Sirca & Adeli, 2018), steel structures (Cha et al., 2018; Kong & Li, 2018; Yeum & Dyke, 2015; Yun et al., 2017), masonry structures (Wang et al., 2018, 2019), structural bolted assemblies (Cha et al., 2016; Pan & Yang, 2021; Ramana et al., 2019). In general, compared to non‐DL‐based methods, DL‐based methods provide higher accuracy and are much more robust against background noise in image classification, object detection, and semantic segmentation.…”