TOWARDS AN INTEGRATED DESIGN METHODOLOGY FOR H-BIM

RGAM: A novel network architecture for 3D point cloud semantic segmentation in indoor scenes

Knowledge about heritage buildings and structures is interesting for a wide variety of stakeholders,
\nranging from architects to operators or the public administration. Such knowledge includes a large
\nvariety of physical and functional characteristics of the building. Laser scanning allows efficient and
\naccurate 3D digitalization of heritage sites and subsequent data processing towards the creation of
\ngeometrically and semantically rich models.
\nParameterization and semantic enrichment of heritage building representations towards the creation
\nof Heritage Building Information Models (HBIM) generally involves the use of point clouds
\nand images as templates for manual mapping procedures in commercial software. Indeed, the information
\nto include in BIM models depends on the requirements of the application it is intended to
\nserve. In view of that, the applications presented in this chapter pertain to automated techniques
\nthat were implemented to parameterize point clouds towards models suitable for energy analysis
\npurposes.
\nThe challenge in automating the reconstruction of heritage buildings is to deal with their geometrical
\ncomplexity and irregularity, meaning that the methodologies selected should be robust and
\nefficient under these conditions. The resulting 3D semantically rich model enhances the knowledge
\nof the heritage building, complementing other representations of the facility such as point clouds or
\nhandmade HBIM. The procedure is implemented and validated in a real case study: the Ducal Palace
\nin Guimarães (Portugal).

Improving performance of deep learning models for 3D point cloud semantic segmentation via attention mechanisms

The present study uses a bibliometric and systematic literature review (SLR) to examine the use of Building Information Modeling (BIM), the Internet of Things (IoT), and Digital Twins (DT) in the construction industry. The network visualization and other approaches based on the Web of Science (WOS) database and the patterns of research interactions were explored in 1879 academic publications using co-occurrence and co-citation investigations. Significant publications, conferences, influential authors, countries, organizations, and funding agencies have been recognized. Our study demonstrates that BIM, IoT, and DT in construction, Heritage BIM (HBIM), Smart Contracts, BIM, and Ontology, and VR and AR in BIM and DT are the main study themes. Finally, several prospective areas for future study are identified, including BIM and Metaverse technology, BIM and Artificial Intelligence (AI), Metaheuristic algorithms for optimization purposes in BIM, and the Circular Economy with BIM and IoT.

Unsupervised domain adaptation (UDA) is increasingly used for 3D point cloud semantic segmentation tasks due to its ability to address the issue of missing labels for new domains. However, most existing unsupervised domain adaptation methods focus only on uni‐modal data and are rarely applied to multi‐modal data. Therefore, we propose a cross‐modal UDA on multi‐modal datasets that contain 3D point clouds and 2D images for 3D Semantic Segmentation. Specifically, we first propose a Dual discriminator‐based Domain Adaptation (Dd‐bDA) module to enhance the adaptability of different domains. Second, given that the robustness of depth information to domain shifts can provide more details for semantic segmentation, we further employ a Dense depth Feature Fusion (DdFF) module to extract image features with rich depth cues. We evaluate our model in four unsupervised domain adaptation scenarios, i.e., dataset‐to‐dataset (A2D2 → SemanticKITTI), Day‐to‐Night, country‐to‐country (USA → Singapore), and synthetic‐to‐real (VirtualKITTI → SemanticKITTI). In all settings, the experimental results achieve significant improvements and surpass state‐of‐the‐art models.

Aiming at the problem that the relationship between geometric features and semantic features is ignored in the point cloud data downsampling process, which leads to inaccurate segmentation of object boundaries and structural details, this paper proposes a 3D point cloud semantic segmentation network based on multi-scale dense nested type. Firstly, a dense nested network architecture is constructed by nesting multiple multi-scale feature fusion modules to fuse multi-scale features of different directions between encoder-decoder paths, so as to effectively propagate local ge-ometric context information and enhance the ability of cross-scale information interaction. Secondly, a local feature aggregation unit is constructed in the multi-scale feature fusion module, which strengthens the structural awareness within the local point set based on graph convolution and attention mechanism, and promotes the complementarity of local geometric features and abstract semantic information. Then, the cross-layer multi-loss supervision module is combined to further optimize the multi-scale feature propagation, which makes the network training more stable and improves the point cloud segmentation accuracy. Finally, this paper verified the proposed network on the S3DIS dataset. The experimental results show that the proposed network has the mean intersection over Union of 71.2% and the overall accuracy of 88.7%, which is 1.2 and 0.7 percentage points higher than RandLA-Net, respectively, which proves that the proposed network can effectively improve the accuracy of 3D point cloud semantic segmentation.

The construction industry has been adopting the new Building Information Modelling (BIM) methodology as the main approach of working, covering all sectors, including design, construction and maintenance. Construction enterprises have been demanding the University to carry out professional courses in order to be able to update their professionals with this new methodology. Recently, a design and construction office requested a one-day BIM training course, to increasing the BIM knowledge of the employees. The short course was properly designed to present the basic concepts, to experience the handling of the BIM modelling software most used, and to disseminate the large range of BIM applicability. The proposed professional action covers the areas of construction (conflict analysis, planning, and material quantity), structures (interoperability, analyses, and the transfer of information between software types), and the recent heritage building information modelling (HBIM) perspective. For it, the applied didactic procedure was based on the presentation of case studies related to situations of conflict between disciplines, of interoperability problems, and of rehabilitation of old buildings. This training involved all the company's employees, composed of designers, architects, civil engineers and economists. The difficulties found in the course are mainly due to the heterogeneity of the attendees, who have different interests and specific familiarity perspectives. The course was held with the objective of initiating professionals in the new topic and to acquiring knowledge in BIM adoption, in a way they can apply in their company. The participants followed the course with great interest and satisfaction, formulating several questions contextualized in the particular field of expertise of each professional. The course aims to contribute to the dissemination of the potential of BIM in the design, construction, and refurbishment of historical buildings. The course was evaluated in its generality and topic taught, resulting in a positive grade.

3D point cloud semantic segmentation has attracted wide attention with its extensive applications in autonomous driving, AR/VR, and robot sensing fields. However, in existing methods, each point in the segmentation results is predicted independently from each other. This property causes the non-contiguity of label sets in three-dimensional space and produces many noisy label points, which hinders the improvement of segmentation accuracy. To address this problem, we first extend adversarial learning to this task and propose a novel framework Attention Adversarial Networks (AttAN). With high-order correlations in label sets learned from the adversarial learning, segmentation network can predict labels closer to the real ones and correct noisy results. Moreover, we design an additive attention block for the segmentation network, which is used to automatically focus on regions critical to the segmentation task by learning the correlation between multi-scale features. Adversarial learning, which explores the underlying relationship between labels in high-dimensional space, opens up a new way in 3D point cloud semantic segmentation. Experimental results on ScanNet and S3DIS datasets show that this framework effectively improves the segmentation quality and outperforms other state-of-the-art methods.

Building Information Modelling (BIM) is a particularly suitable methodology to implement a multidisciplinary approach in dealing problems that involve a large amount of data. This condition is typical of cultural heritage buildings, that are intrinsically complex (due to construction phases, variability of materials, unusual architectural solutions). However, the complexity of existent buildings makes very difficult to reduce them to repetitive schemes, useful to set an ordinary BIM model. For this reason, currently the application of BIM to the cultural heritage field (properly called HBIM - Heritage Building Information Modelling) is still not sufficiently tuned and need further reflection and development. In this paper case studies are presented in which the processes of knowledge, modelling, and conservation are applied to some Italian cultural heritage buildings of small dimension. The discussed applications are mainly referred to seismic retrofitting interventions, that are particularly sensitive because they require a holistic approach involving the knowledge of the construction phases, construction techniques, production methods, and the ageing processes of the historical materials.

Semantic segmentation of 3D point cloud data is essential for enhanced high-level perception in autonomous platforms. Furthermore, given the increasing deployment of LiDAR sensors onboard of cars and drones, a special emphasis is also placed on non-computationally intensive algorithms that operate on mobile GPUs. Previous efficient state-of-the-art methods relied on 2D spherical projection of point clouds as input for 2D fully convolutional neural networks to balance the accuracy-speed trade-off. This paper introduces a novel approach for 3D point cloud semantic segmentation that exploits multiple projections of the point cloud to mitigate the loss of information inherent in single projection methods. Our Multi-Projection Fusion (MPF) framework analyzes spherical and bird's-eye view projections using two separate highly-efficient 2D fully convolutional models then combines the segmentation results of both views. The proposed framework is validated on the SemanticKITTI dataset where it achieved a mIoU of 55.5 which is higher than state-of-the-art projection-based methods RangeNet++ [23] and PolarNet [44] while being 1.6x faster than the former and 3.1x faster than the latter.

Semantic segmentation is an important component in understanding the 3D point cloud scene. Whether we can effectively obtain local and global contextual information from points is of great significance in improving the performance of 3D point cloud semantic segmentation. In this paper, we propose a self-attention feature extraction module: the local transformer structure. By stacking the encoder layer composed of this structure, we can extract local features while preserving global connectivity. The structure can automatically learn each point feature from its neighborhoods and is invariant to different point orders. We designed two unique key matrices, each of which focuses on the feature similarities and geometric structure relationships between the points to generate attention weight matrices. Additionally, the cross-skip selection of neighbors is used to obtain larger receptive fields for each point without increasing the number of calculations required, and can therefore better deal with the junction between multiple objects. When the new network was verified on the S3DIS, the mean intersection over union was 69.1%, and the segmentation accuracies on the complex outdoor scene datasets Semantic3D and SemanticKITTI were 94.3% and 87.8%, respectively, which demonstrate the effectiveness of the proposed methods.

3D point cloud semantic segmentation is a fundamental task for 3D scene understanding. However, most existing pipelines usually use k-NN or ball query operation to form hard neighborhoods, which may cross different semantic objects, resulting low-quality local features. To address this issue, we propose a multi-scale superpoint network that gradually generates multi-scale soft neighborhoods to extract geometric local features, thereby boosting the 3D semantic segmentation performance. Specifically, we present a simple yet efficient superpoint merging module that merge small-scale superpoints to obtain large-scale superpoint by considering the feature similarity of superpoints, so that we can obtain multi-scale geometric features of point clouds. We also develop a superpoint upsampling module that adopt inverse mapping function to propagate multi-scale features from low-resolution point cloud to high-resolution point cloud. By integrating our multi-scale superpoint network into a simple point based semantic segmentation network, our method can obtain SOTA results on S3DIS Area 5 and 6-fold, and competitive results on ScanNet v2.

The use of Building Information Modelling (BIM) for Building Energy Simulation (BES) is challenged by inadequate data integration and unreliable data and geometry conversion. Despite recent efforts to address critical interoperability issues, the effective exchange of data between the two digital environments remains a non-truly fulfilled promise of the digital transformation agenda for the industry. To contribute to the broader community effort in mitigating this deficiency, the present article proposes an alternative method for securing an improved BIM to BES workflow for existing and heritage buildings. The proposed method suggests an alternative way for integrating environmental analysis data to the Heritage BIM model, as well as an easy to develop visual programming process for missing data recovery, through auxiliary exchange operations. The article also presents lessons learned for the geometrical conversion of BIM elements with non-standard and irregular attributes, evaluated through the energy retrofit application of a colonial heritage building. The proposed method demonstrates how visual programming tools and common data exchange formats can supplementary assist existing practises to deal with major interoperability obstacles. The overall findings contribute to current efforts in the field, for the establishment of good practices of using Heritage BIM for energy retrofit architectural interventions.

Over the ages, Egypt has been known for its flourished history with the ability and the power to fascinate mankind. Its great heritage appears in several forms that include monuments, tombs, temples, and documents revealing past events, but until the beginning of the 21st century only a minimum share of the secrets and the mysteries were exposed. Egypt’s monuments are deteriorated and out of the spot and has been left without maintenance and development. Numerous monuments and historical sites are ignored and not even considered as a destination for tourists because of the absence of virtuous marketing and propaganda. The planning and management of cultural heritage buildings through digital 3D models has become a necessity to protect and conserve the existing history. Building Information Modelling BIM is considered a main tool adopted by the Architecture, Engineering and Construction AEC sector for its qualities and capabilities. Due to the complexity of cultural heritage buildings the concept of Heritage Building Information Modelling (HBIM) which concerns the modelling, documentation, and virtualizing the heritage. This research will focus on the use of HBIM technology by implementing augmented reality mobile application to upgrade the quality of branding of historical sites in Egypt. The AR will augment Al-Baron Palace, which is the chosen case study. The methodology adopted is literature review, comparative analysis on international case studies of sites that applied augmented reality, and an experiment followed by a questionnaire answered by users to evaluate the usage of such mobile application. The research will conclude with a framework that will provide guidelines for applying augmented reality technology system in order to enhance and promote Egypt’s heritage.

As a pivotal task in 3D scene understanding, 3D point cloud semantic segmentation is often challenged by the presence of mislabeled data in real-world annotations. Such noisy labels can mislead the segmentation model into learning erroneous features during training, thereby degrading its performance. To address this issue, this paper proposes a robust training framework for 3D semantic segmentation based on noisy label correction. The proposed method does not rely on prior knowledge of point cloud density and is capable of learning robust segmentation models under various noise rates. Our approach consists of two core components: a Density-Adaptive Neighborhood Clustering algorithm that dynamically adjusts the neighborhood radius and minimum point count according to local point density variations, and a Dynamically-Weighted Label Refurbishing mechanism, which rectifies noisy labels at the cluster level based on historical prediction confidence. Extensive experiments on the S3DIS dataset demonstrate that our method outperforms existing baselines and state-of-the-art (SOTA) approaches under symmetric, asymmetric, and mixed noise settings. Notably, it maintains strong robustness even under high noise ratios. Furthermore, our method integrates label structures and stores them in a vectorized format, significantly reducing data transfer overhead between CPU and GPU and mitigating resource wastage. This design makes the approach particularly efficient in resource-constrained environments such as cloud servers, where CPU performance is often a bottleneck. As a result, our method achieves a notable improvement in training efficiency, with each training epoch requiring only approximately one-tenth of the time compared to current SOTA methods.