| 1 | Alawadhi and Yan (2021) | 2021 | Presented a concept for a hybrid system that uses BIM and hyperrealistic rendering for DL to synthesize data sets for training a neural network for object recognition in photographs, with the goal of building object recognition in photographs |
| 2 | Chang et al. (2019) | 2019 | By applying DL and BIM to indoor positioning, we were able to develop a sound-based method |
| 3 | Chen and Demachi (2021) | 2021 | The combination of DL-based object detection and individual detection using geometric relationships analysis resulted in a novel solution for identifying improper use of personal protective equipment |
| 4 | Hou et al. (2021a) | 2021 | Identified relevant gaps, challenges and future work on DL-based applications for safety management in the AEC industry |
| 5 | Kim and Lee (2020) | 2020 | With reference images and a deep-learning model, this paper describes a stochastic approach for identifying and appending interior design style information to a design document |
| 6 | Marzouk and Zaher (2020) | 2020 | DL was used to present a methodology for exploitation of artificial intelligence in facility management using DL |
| 7 | Ma et al. (2020) | 2020 | With DL, a methodology for generating synthetic point clouds from 3D BIM models was demonstrated and discussed |
| 8 | Perez-Perez et al. (2021) | 2021 | Presented an end-to-end DL method, Scan2BIM-NET, for semantically segmenting the structural, architectural and mechanical components present in point cloud data, as well as experimental results demonstrating the method's effectiveness |
| 9 | Wu et al. (2020) | 2020 | DL and NLP techniques were used to develop a method for screening patents related to information and communications technology (ICT) in construction |
| 10 | Xue and Zhang (2021) | 2021 | It was proposed that a part-of-speech (POS) tagger tailored to building codes be developed, which would be empowered by DL and transformative rules |
| 11 | Zhong et al. (2020a) | 2020 | In order to improve the efficiency and effectiveness of retrieving queries pertaining to building regulations, a robust end-to-end methodology was developed that integrates information retrieval with a DL model of natural language processing (NLP) (NLP) |