Cumincad : CUMINCAD Papers : Search Results

CumInCAD is a Cumulative Index about publications in Computer Aided Architectural Design
supported by the sibling associations ACADIA, CAADRIA, eCAADe, SIGraDi, ASCAAD and CAAD futures

PDF papers
References

Hits 1 to 20 of 332

Reformat results as: short short into frame detailed detailed into frame

_id	ecaade2024_114
id	ecaade2024_114
authors	Su, Xinyu; Liu, Zidong; Yang, Mingzhuo; Koehler, Daniel
year	2024
title	ZoeLength: Framework for indoor measurement from a single interior image for the popularization of AI interior design
doi	https://doi.org/10.52842/conf.ecaade.2024.2.413
source	Kontovourkis, O, Phocas, MC and Wurzer, G (eds.), Data-Driven Intelligence - Proceedings of the 42nd Conference on Education and Research in Computer Aided Architectural Design in Europe (eCAADe 2024), Nicosia, 11-13 September 2024, Volume 2, pp. 413–422
summary	Applying Artificial Intelligence Generated Content (AIGC) to interior design makes it possible for anyone to take on the role of a designer. Size is crucial to interior design, and distance measurement has become an essential component allowing to integration of image generation in industrial supply chain processes. This paper explores a new framework for indoor measurement based on a single interior image. Without any reference and camera calibration, our method ZoeLength can estimate the target size by taking a photo with the simplest mobile device. We achieved this by constructing a simplified camera model, incorporating cutting-edge depth estimation technology, ZoeDepth and Depth Anything, and object detection technology, Grounding DINO. To increase the accuracy of measurement, we trained a depth estimation model specifically for indoor scenes using our own collected dataset. Experimental results from multiple aspects demonstrate the reliability and validity of the proposed method and its application value to real-world scenarios.
keywords	AIGC Interior Design, Indoor Measurement, Single-Image Measurement, Depth Estimation, Object Detection
series	eCAADe
email
last changed	2024/11/17 22:05

_id	caadria2024_68
id	caadria2024_68
authors	Chen, Changyu, Yu, Hanting and Guo, Yuhan
year	2024
title	Perception and Reality: Urban Green Space Analysis Using Language Model-Based Social Media Insights. A Case Study within Shanghai’s Inner Ring
doi	https://doi.org/10.52842/conf.caadria.2024.2.119
source	Nicole Gardner, Christiane M. Herr, Likai Wang, Hirano Toshiki, Sumbul Ahmad Khan (eds.), ACCELERATED DESIGN - Proceedings of the 29th CAADRIA Conference, Singapore, 20-26 April 2024, Volume 2, pp. 119–128
summary	As urbanization accelerates, urban green spaces play an increasingly integral role in daily life. The development of large language models (LLM) provides a practical method for assessing how people perceive green spaces. This study employs prompt-turning techniques to analyze social media data, aiming to uncover public sentiments and correlate them with the actual conditions of urban green spaces. By categorizing evaluative dimensions — location, landscape quality, facility levels, and management — diverse areas of public focus are revealed. We utilize both dictionary-based and language model-based methods for the analysis of overall emotional perception and dimensional emotional perception. Classification of sentiments into positive, neutral, and negative categories enhances our understanding of the public's general emotional inclinations. Using various spatial analysis techniques, the study delves into the current conditions of green spaces across these evaluative dimensions. In conclusion, a correlation analysis exposes patterns and disparities in these evaluative dimensions, providing valuable insights into understanding public emotional tendencies and offering effective recommendations from the perspective of public perception.
keywords	Urban green spaces, Social media analysis, Public Sentiment, Spatial analysis, Prompt-turning
series	CAADRIA
email
last changed	2024/11/17 22:05

_id	ecaade2024_66
id	ecaade2024_66
authors	Jabi, Wassim; Li, Yang
year	2024
title	Graph Neural Networks for Node Classification and Attribute Allocation in Architectural BIM
doi	https://doi.org/10.52842/conf.ecaade.2024.1.675
source	Kontovourkis, O, Phocas, MC and Wurzer, G (eds.), Data-Driven Intelligence - Proceedings of the 42nd Conference on Education and Research in Computer Aided Architectural Design in Europe (eCAADe 2024), Nicosia, 11-13 September 2024, Volume 1, pp. 675–684
summary	Building Information Modelling (BIM) marks a notable shift in architectural design, extending beyond simple digital reproductions by capturing the spatial, physical, and operational characteristics of structures. Unfortunately, these representations are often complex in nature and difficult to inspect, analyze, and understand which can lead to errors and omissions during model construction. This research aims to leverage graph machine learning systems, utilizing learned datasets, to detect and rectify these issues, improving model quality and minimizing costly mistakes. To illustrate the application of graph neural networks in this domain, this paper applied a graph-based geometric and topological editor coupled with a graph neural network to a real-world dataset of residential building complexes. The developed workflow operates by converting traditional architectural floor plans into graph-structured data, enabling precise node classification predictions. The paper details the overall workflow, data preparation and conversion, hyperparameter optimization and experimental results. Comparing the performance of various graph neural network models has validated the efficiency of the chosen prediction model in processing and analyzing architectural floor plans, achieving an overall accuracy rate of approximately 95%. The paper concludes with a discussion of the potential and limitations of graph-based machine learning methodologies within the architectural domain and an outline of future work plans.
keywords	Topology, Artificial Intelligence, Machine Learning, Graph Neural Network, Node Classification, Floor Plans
series	eCAADe
email
last changed	2024/11/17 22:05

_id	ecaade2024_242
id	ecaade2024_242
authors	Koh, Immanuel; Saw, Man Lin
year	2024
title	Architectural Dramaturgy: A total and endless theatre with multimodal artificial intelligence
doi	https://doi.org/10.52842/conf.ecaade.2024.1.559
source	Kontovourkis, O, Phocas, MC and Wurzer, G (eds.), Data-Driven Intelligence - Proceedings of the 42nd Conference on Education and Research in Computer Aided Architectural Design in Europe (eCAADe 2024), Nicosia, 11-13 September 2024, Volume 1, pp. 559–564
summary	Close to a century ago, Walter Gropius proposed the highly influential but unrealised project ‘Total Theatre’ (1927). The project captured Gropius’ adapted formulation of ‘Gesamtkunstwerk’(‘total work of art’) as ‘Total Design’ within the Bauhaus, and more specifically, in relation to ‘The Theatre of the Bauhaus’ (1924). The ‘Total Theatre’ was Gropius’ attempt to dynamically reconfigure the previously static relationship between actors and audience into one that is participatory and filmic in spatial layout. Around the same time in history, shortly after designing the set for Eugene O’Neill’s ‘The Emperor Jones’ in 1924, Frederick Kiesler began to conceptualise his ‘Endless Theatre’ (1926). In the former, Kiesler created a constructivist stage space with moving walls/floors/ceilings animating in-sync with the play’s narrative and producing cinematographic effects similar to those found in films. For both Gropius and Kiesler, the theatre is a fertile ground for incorporating the latest technologies to prototype a ‘totally’ multimodal and ‘endlessly’ generative form of architecture. Against the backdrop of such past intellectual efforts in architecture, and that of today’s rapidly foregrounding of Large Language Models (LLMs) and Large Multimodal Models (LMMs) in artificial intelligence (AI) within the architecture discipline, the paper posits that it is timely to revisit this conceptual cross-fertilisation of theatre and architecture – architectural dramaturgy. In doing so, the research aims to extend Gropius’s and Kiesler’s concept of the ‘total’ and the ‘endless’ respectively through the lens of the ‘computational’. The first experiment uses different datasets from existing established theatre creators to train and prompt-engineer relevant AI models, namely a text-to-text model for playwright Arthur Miller, a text-to-image model for director Ivo van Hove, and a text-to-audio model for composer Stephen Sondheim. In improving the coherence of the results, the second experiment replaces unimodality with multimodality leveraging a single source of video data (the poignant “To be or not to be” soliloquy from Shakespeare’s Hamlet) to formulate a human-in-the-loop interpretative framework by utilising a combination of text-image-to-image model and text-image-to-video model, and further postprocessed with image inpainting model and image-to-3D model. With the deliberate bracketing of site, programmes and other specificities typical of an architecture project, the research demonstrates how concepts borrowed from theatre when layered with multimodal AI could extend the discipline’s longstanding conception of a total and endless architecture.
keywords	Theatre, Deep Learning, Large Multimodal Models, Kiesler, Gropius
series	eCAADe
email
last changed	2024/11/17 22:05

_id	caadria2024_166
id	caadria2024_166
authors	Li, Jinmin, Luo, Yilu, Lu, Shuai, Zhang, Jingyun, Wang, Jun, Guo, Rizen and Wang, ShaoMing
year	2024
title	ChatDesign: Bootstrapping Generative Floor Plan Design With Pre-trained Large Language Models
doi	https://doi.org/10.52842/conf.caadria.2024.1.099
source	Nicole Gardner, Christiane M. Herr, Likai Wang, Hirano Toshiki, Sumbul Ahmad Khan (eds.), ACCELERATED DESIGN - Proceedings of the 29th CAADRIA Conference, Singapore, 20-26 April 2024, Volume 1, pp. 99–108
summary	Large language models (LLMs) have achieved remarkable success in various domains, revolutionizing tasks such as language translation, text generation, and question-answering. However, generating floor plan designs poses a unique challenge that demands the fulfilment of intricate spatial and relational constraints. In this paper, we propose ChatDesign, an innovative approach that leverages the power of pre-trained LLMs to generate floor plan designs from natural language descriptions, while incorporating iterative modifications based on user interaction. By processing user input text through a pre-trained LLM and utilizing a decoder, we can generate regression parameters and floor plans that are precisely tailored to satisfy the specific needs of the user. Our approach incorporates an iterative refinement process, optimizing the model output by considering the input text and previous results. Throughout these interactions, we employ many strategic techniques to ensure the generated design images align precisely with the user's requirements. The proposed approach is extensively evaluated through rigorous experiments, including user studies, demonstrating its feasibility and efficacy. The empirical results consistently demonstrate the superiority of our method over existing approaches, showcasing its ability to generate floor plans that rival those created by human designer. Our code will be available at https://github.com/THU-Kingmin/ChatDesign.
keywords	floor plan generation, large language models, user interactions, automatic design, deep learning, pre-train models
series	CAADRIA
email
last changed	2024/11/17 22:05

_id	ecaade2024_297
id	ecaade2024_297
authors	Massafra, Angelo; Coraglia, Ugo Maria; Predari, Giorgia; Gulli, Riccardo
year	2024
title	Building Information Model Analysis Through Large Language Models and Knowledge Graphs
doi	https://doi.org/10.52842/conf.ecaade.2024.1.685
source	Kontovourkis, O, Phocas, MC and Wurzer, G (eds.), Data-Driven Intelligence - Proceedings of the 42nd Conference on Education and Research in Computer Aided Architectural Design in Europe (eCAADe 2024), Nicosia, 11-13 September 2024, Volume 1, pp. 685–694
summary	The advent of Large Language Models (LLMs) seems to mark a break between past and present in the methods of structuring knowledge, making it possible today to transfer this capability to machines even in a sector like AECO, always been information-intensive but resistant to technological transition. In terms of knowledge, the most established paradigm has been Building Information Modelling (BIM), with IFC functioning as the main schema for standardizing the industry's information. Added to this are knowledge graphs that, emerging with semantic web technologies, allow storing knowledge in structures consisting of nodes and edges with semantic meanings. Nevertheless, a barrier to the widespread adoption of BIM is its accessibility. Querying BIM models is often limited for stakeholders without digital skills, who may struggle to access the vast amount of information stored in these complex informative models. In an attempt to outline one of the possible uses of LLMs in BIM, this research proposes a method for querying BIM models through textual prompts aimed at analyzing a selected case study. In the workflow, a BIM model is first realized. Then, data is integrated into a knowledge graph. Next, ChatGPT's LLMs are used to activate query functions for the analysis of the graph. The results of the queries are displayed in a user-friendly graphical user interface. The study's outcomes offer insights for researchers and industry professionals, highlighting emerging research potentials for LLMs in the field.
keywords	Building Information Modeling, Large Language Models, Natural Language Processing, Knowledge Graphs
series	eCAADe
email
last changed	2024/11/17 22:05

_id	caadria2024_156
id	caadria2024_156
authors	Middlemiss, Aleisha, Welch, Chris, Innes, Daniel, Austin, Matthew and Morgan, Tristan
year	2024
title	Enabling Design Partnerships Through Parametric Design Visualisation: A Case Study of Visual Communication Techniques Implemented on an Infrastructure Design Project Where Partnership With Indigenous Communities was Integral to Success
doi	https://doi.org/10.52842/conf.caadria.2024.2.211
source	Nicole Gardner, Christiane M. Herr, Likai Wang, Hirano Toshiki, Sumbul Ahmad Khan (eds.), ACCELERATED DESIGN - Proceedings of the 29th CAADRIA Conference, Singapore, 20-26 April 2024, Volume 2, pp. 211–220
summary	Where there is a commitment authentically engage indigenous communities in the design decision process, the relationships and communication methods that are used are an integral part of the design project. In an Aotearoa New Zealand context, mana whenua (the name given to the local indigenous group with historical rights to the land) are engaged at various levels with the Crown on infrastructure projects to achieve mutually beneficial outcomes in alignment with Te Tiriti o Waitangi. Infrastructure projects are communicated using specialist language and drawings, which is often unfamiliar to non-professionals. This is a learning opportunity for non-professional stakeholders and a pre-requisite to holistically understand and therefore meaningfully contribute to a project. This paper describes a case study of the evolving application of computational design as an additive communication technique for consultation sessions with mana whenua. This paper compares 2D documentation, pre-rendered 3D models, and interactive parametric modelling approaches. A major insight gained through this comparison of approaches was that enabling mana whenua to fully engage with a 3D model through interactive parametric modelling was effective in deepening understanding of the project. This deeper understanding of the project and its context could enable them to utilise their tacit knowledge of the land to make important contributions crucial to the project and ensure the project is in alignment with mana whenua values and aspirations.
keywords	Infrastructure, Collaboration, Computational Design, Integrated Design
series	CAADRIA
email
last changed	2024/11/17 22:05

_id	ecaade2024_117
id	ecaade2024_117
authors	Su, Xinyu; Luo, Jianhe; Liu, Zidong; Yan, Gaoliang
year	2024
title	Text to Terminal: A framework for generating airport terminal layout with large-scale language-image models
doi	https://doi.org/10.52842/conf.ecaade.2024.1.469
source	Kontovourkis, O, Phocas, MC and Wurzer, G (eds.), Data-Driven Intelligence - Proceedings of the 42nd Conference on Education and Research in Computer Aided Architectural Design in Europe (eCAADe 2024), Nicosia, 11-13 September 2024, Volume 1, pp. 469–478
summary	Large-scale language-image (LLI) models present novel opportunities for architectural design by facilitating its multimodal process via text-image interactions. However, the inherent two-dimensionality of their outputs restricts their utility in architectural practice. Airport terminals, characterized by their flexibility and patterned forms, with most of the design operations occurring at the level of master plan, indicating a promising application area for LLI models. We propose a workflow that, in the early design phase, employs a fine-tuned Stable Diffusion model to generate terminal design solutions from textual descriptions and a site image, followed by a quantitative evaluation from an architectural expert's viewpoint. We created our dataset by collecting satellite images of 295 airport terminals worldwide and annotating them in terms of size and form. Using Terminal 2 of Zhengzhou Xinzheng International Airport as a case study, we scored the original and generated solutions on three airside evaluation metrics, verifying the validity of the proposed method. Our study bridges image generation and expert architectural design assessments, providing valuable insights into the practical application of LLI models in architectural practice and introducing a new method for the intelligent design of large-scale public buildings.
keywords	Multimodal Machine Learning, Diffusion Model, Text-to-Architecture, Airport Terminal Configuration Design, Post-Generation Evaluation
series	eCAADe
email
last changed	2024/11/17 22:05

_id	ecaade2024_153
id	ecaade2024_153
authors	Tsurunaga, Shinya; Fukuda, Tomohiro; Yabuki, Nobuyoshi
year	2024
title	Enhanced Landscape Visualization of Post-Structure Removal: Integrating 3D reconstruction techniques and diffusion models through machine learning
doi	https://doi.org/10.52842/conf.ecaade.2024.1.549
source	Kontovourkis, O, Phocas, MC and Wurzer, G (eds.), Data-Driven Intelligence - Proceedings of the 42nd Conference on Education and Research in Computer Aided Architectural Design in Europe (eCAADe 2024), Nicosia, 11-13 September 2024, Volume 1, pp. 549–558
summary	In urban redevelopment, demolition of existing buildings often occur and landscape assessment plays an important role in avoiding various environmental issues. Both residents and professionals should be involved to create a virtual three-dimensional (3D) space after demolition, which would enable even non-experts to understand the future landscape. Research efforts aimed at creating virtual 3D spaces by removing unnecessary objects utilize techniques such as neural radiance fields (NeRF). These techniques reconstruct spaces into virtual 3D spaces from RGB images by removing redundant objects. However, a challenge arises from the low-quality images generated from the resultant space. Additionally, methods for reconstructing 3D images face limitations in acquiring images of portions previously obscured by structures slated for demolition. This often leads to numerous artifacts in 3D reconstruction after structure removal, which hinders accurate space construction. This study proposes a system that integrates 3D Gaussian splatting, capable of high-quality 3D reconstruction through machine learning, and image completion processing using a diffusion model. This integration aims to reduce the impact of artifacts in 3D reconstruction after building removal in complex and large-scale urban areas. This will contribute to the intuitive understanding and decision-making of non-experts, such as residents, in future landscape assessments after building removal.
keywords	3D reconstruction, diffusion model, landscape visualization, view synthesis, real-time rendering, 3D Gaussian splatting
series	eCAADe
email
last changed	2024/11/17 22:05

_id	caadria2024_58
id	caadria2024_58
authors	Zhuang, Junling, Li, Guanhong, Xu, Hang, Xu, Jintu and Tian, Runjia
year	2024
title	Text-to-City: Controllable 3D Urban Block Generation With Latent Diffusion Model
doi	https://doi.org/10.52842/conf.caadria.2024.2.169
source	Nicole Gardner, Christiane M. Herr, Likai Wang, Hirano Toshiki, Sumbul Ahmad Khan (eds.), ACCELERATED DESIGN - Proceedings of the 29th CAADRIA Conference, Singapore, 20-26 April 2024, Volume 2, pp. 169–178
summary	The rise of deep learning has introduced novel computational tools for urban block design. Many researchers have explored generative urban block design using either rule-based or deep learning methods. However, these methods often fall short in adequately capturing morphological features and essential design indicators like building density. Latent diffusion models, particularly in the context of urban design, offer a groundbreaking solution. These models can generate cityscapes directly from text descriptions, incorporating a wide array of design indicators. This paper introduces a novel workflow that utilizes Stable Diffusion, a state-of-the-art latent diffusion model, to generate 3D urban environments. The process involves reconstructing 3D urban block models from generated depth images, employing a systematic depth-to-height mapping technique. Additionally, the paper explores the extrapolation between various urban morphological characteristics, aiming to generate novel urban forms that transcend existing city models. This innovative approach not only facilitates the accurate generation of urban blocks with specific morphological characteristics and design metrics, such as building density, but also demonstrates its versatility through application to three distinct cities. This methodology, tested on select cities, holds potential for broader range of urban environments and more design indicators, setting the stage for future computational urban design research.
keywords	deep learning, generative design, latent diffusion model, urban block morphology, artificial intelligence
series	CAADRIA
email
last changed	2024/11/17 22:05

_id	ecaade2024_201
id	ecaade2024_201
authors	Hashizume, Keiji; Fukuda, Tomohiro; Yabuki, Nobuyoshi
year	2024
title	A Surface Modeling Method for Indoor Spaces from 3D Point Cloud Reconstructed by 3D Gaussian Splatting
doi	https://doi.org/10.52842/conf.ecaade.2024.1.695
source	Kontovourkis, O, Phocas, MC and Wurzer, G (eds.), Data-Driven Intelligence - Proceedings of the 42nd Conference on Education and Research in Computer Aided Architectural Design in Europe (eCAADe 2024), Nicosia, 11-13 September 2024, Volume 1, pp. 695–704
summary	Building information modeling (BIM) is becoming increasingly important in architectural projects, and the implementation of BIM in new construction projects is progressing. On the other hand, many existing buildings do not have BIM data, so it is necessary to create it from scratch. A common method for converting existing buildings to BIM is scan-to-BIM, using techniques such as laser scanning or photogrammetry. However, laser scanning provides accurate point cloud data but requires expensive equipment, while photogrammetry is generally cost-effective but has lower accuracy point cloud data. Another approach for creating BIM from 2D images is to use neural radiance fields (NeRF). However, NeRF faces challenges in terms of data accuracy and processing speed when dealing with large or complex scenes. In contrast, 3D Gaussian Splatting is an emerging computer vision technology that uses machine learning to reconstruct 3D scenes from 2D images faster than NeRF, with comparable or better quality. Therefore, this study proposes a method to create surface models consisting of floors, walls, and ceilings as a preliminary step to creating BIM data for existing indoor spaces using 3D Gaussian Splatting. First, point cloud is generated using 3D Gaussian Splatting, followed by noise reduction. The point cloud is then classified based on height. Subsequently, processing such as extraction of boundary primitives from the point cloud of the floor and classification of feature points are performed to estimate the shape of the floor. Finally, ceilings and walls are created based on height and floor shape. The results of validation confirm an error of between 0.01m and 0.5m in the generated surface models. This study proposes a novel attempt to create 3D models using 3D Gaussian Splatting, contributing to the generation of BIM data for existing buildings.
keywords	Point Cloud, 3D Gaussian Splatting, Scan2BIM, Surface Modeling, Indoor 3D Reconstruction
series	eCAADe
email
last changed	2024/11/17 22:05

_id	caadria2024_114
id	caadria2024_114
authors	Yin, Jun, Xu, Pengjian, Gao, Wen, Zeng, Pengyu and Lu, Shuai
year	2024
title	Drag2build: Interactive Point-Based Manipulation of 3D Architectural Point Clouds Generated From a Single Image
doi	https://doi.org/10.52842/conf.caadria.2024.1.169
source	Nicole Gardner, Christiane M. Herr, Likai Wang, Hirano Toshiki, Sumbul Ahmad Khan (eds.), ACCELERATED DESIGN - Proceedings of the 29th CAADRIA Conference, Singapore, 20-26 April 2024, Volume 1, pp. 169–178
summary	At present, 3D reconstruction from images has made notable advancements in simple, small-scale scenes, but faces significant challenges in intricate, expansive architectural scenes. Focusing on the early stage of design stage, we present Drag2Build, a tool for converting images into point clouds for 3D reconstruction and modification in detailed architectural contexts. Our first step involved the creation of ArchiNet, a specialized 3D reconstruction dataset dedicated to elaborate architectural scenes. Next, we developed a 3D reconstruction approach using a conditional denoising diffusion model, enhanced by incorporating a model for segmenting objects, thereby improving segmentation and identification in complex scenes. Additionally, our system features an interactive component that allows for immediate modification of 2D images via an easy drag-and-drop action, synchronously updating 3D architectural point clouds. The performance of Drag2Build in 3D reconstruction precision was assessed and benchmarked against mainstream methods using ArchiNet. The experiments showed that our approach is capable of producing high-quality 3D point clouds, facilitating swift editing and efficient handling of intricate backgrounds.
keywords	3D building Generation, Diffusion Model, Single Image Reconstruction, DragDiffusion
series	CAADRIA
email
last changed	2024/11/17 22:05

_id	architectural_intelligence2024_25
id	architectural_intelligence2024_25
authors	Ana Goidea, Mariana Popescu, Anton Tetov Johansson & David Andréen
year	2024
title	Algorithmic modeling of functionally graded metamaterials in 3D printed building envelopes
doi	https://doi.org/https://doi.org/10.1007/s44223-024-00068-x
source	Architectural Intelligence Journal
summary	Recent development of powder-bed additive manufacturing promises to enable the production of architectural structures that combine high resolution and articulation with economies of scale. These capabilities can potentially be used for functionally graded metamaterials as part of the building envelope and structure, paving the way for new functionalities and performances. However, designing such multifunctional structures requires new design and modelling strategies to control, understand, and generate complex geometries and their transcalar interdependencies. The work presented here demonstrates a modeling framework that can unite multiple generative and organizational algorithms to create a unified, 3D printable building element that integrates a range of functional requirements. Our methods are based on an understanding of stigmergic principles for self-organization and developed to allow for a wide range of application scenarios and design intents. The framework is structured around a composite modeling environment based on a combination of volumetric modeling and particle-spring systems, and is developed to negotiate the large scalar range necessary for such applications. We present here a prototype demonstrator designed using this framework: Meristem Wall, a functionally integrated building envelope fabricated through a combination of powder bed 3D printing and CNC knitting.
series	Architectural Intelligence
email
last changed	2025/01/09 15:05

_id	caadria2024_125
id	caadria2024_125
authors	Bai, Zishen and Zhang, Chen
year	2024
title	A Discriminator with Deep Learning (ResNet-LST) for Evaluating the Impact of Urban Morphological Indicators on Urban Land Surface Temperatures
doi	https://doi.org/10.52842/conf.caadria.2024.2.009
source	Nicole Gardner, Christiane M. Herr, Likai Wang, Hirano Toshiki, Sumbul Ahmad Khan (eds.), ACCELERATED DESIGN - Proceedings of the 29th CAADRIA Conference, Singapore, 20-26 April 2024, Volume 2, pp. 9–18
summary	With global urban temperatures increasing yearly, city residents face serious health threats. Studies have shown a significant correlation between urban morphology and urban Land Surface Temperature (LST). However, the research findings of urban morphology and LST are frequently inconsistent or contradictory, which still need further exploration. The study proposes a ResNet-LST, which uses a Convolutional Neural Network (CNN) image classification model to link urban morphology with urban LST. The research aims to develop an effective real-time assessment tool for architects/urban designers to evaluate the impact of urban morphology indicators on urban LST. The ResNet-LST model went through 950 iterations with an overall accuracy of 79.48%. As a contribution, the study demonstrates that the surface temperature of each city region can reflect its contribution to the global temperature. Furthermore, the research results demonstrate the powerful flexibility in design decision-making for fast interaction using data-driven deep learning techniques. Designers will no longer need to pay high costs for LST simulations; image-based assessment models could give prompt feedback by recognizing 2D graphics of design proposals.
keywords	LST, Urban form Indicators, Image Classification, Convolutional Neural Network, Open Urban Data
series	CAADRIA
email
last changed	2024/11/17 22:05

_id	ecaade2024_338
id	ecaade2024_338
authors	Blahut, Sarah; Faizi, Harun; Harnoncourt-Fuchs, Marie-Therese
year	2024
title	Hybrid Workflow for Feedback using Extended Reality and 3D Scanning Systems
doi	https://doi.org/10.52842/conf.ecaade.2024.2.675
source	Kontovourkis, O, Phocas, MC and Wurzer, G (eds.), Data-Driven Intelligence - Proceedings of the 42nd Conference on Education and Research in Computer Aided Architectural Design in Europe (eCAADe 2024), Nicosia, 11-13 September 2024, Volume 2, pp. 675–684
summary	The research presented questions standard representational tools used in methods for project analysis and presentation in teaching. The experiment presented tests a hybrid workflow relying on extended reality and 3D scanning systems, creating a feedback loop in each step, expanding on current research. The work aims to leverage the potential in using advanced technology in a series of case studies, carried out by a group of 25 architecture students in three steps. The goal was for students to learn essential skillsets for Rhino 3D, 3D scanning methods, and virtual (VR) and mixed reality (MR) systems for analyzing complex architectural structures. The experiment was conducted using hybrid workflows, in groups of two and three, to enhance a research task of analyzing the design and construction of a timber pavilion. The three steps in the analysis are based on the initial reconstruction of the pavilion in Rhino 3D, establishing key aspects of the structure as 3D model data. In this initial step, MR systems are implemented to test, review, and adjust the digital 3D model in an interactive immersive environment. In the following step, the assembly logic identified in the structure is tested by using MR systems developed for fabricating a scaled physical model. In the final step, the physical model built in the previous step is 3D scanned. The resulting 3D model data is compared with the initial Rhino 3D model data assisted with VR systems and are evaluated for legibility, accuracy in execution, and degrees of fidelity. The approach offers a hybrid learning context that established feedback loops within interactive, immersive environments. This also enhanced the learning task and production of complex spatial structures and 3D data, also highlighting areas for improvement. The hybrid workflow has provided other opportunities for future work through the integration of other advanced technologies and further testing in scalable iterations within in the design and building process.
keywords	Virtual Reality, Mixed Reality, Representation tools, HoloLens2, OculusQuest2, Digital and Analog, Augmented Reality, 3D scanning, Comparative Analysis, Feedback Loop, Project Analysis, Presentation Methods, Teaching, Hybrid Workflow
series	eCAADe
email
last changed	2024/11/17 22:05

_id	ecaade2024_85
id	ecaade2024_85
authors	Casakin, Hernan; Sopher, Hadas; Anidjar, Or H.; Gero, John S.
year	2024
title	A Data-Driven NLP Approach to Analyzing Framing and Reframing in Design Protocols
doi	https://doi.org/10.52842/conf.ecaade.2024.2.547
source	Kontovourkis, O, Phocas, MC and Wurzer, G (eds.), Data-Driven Intelligence - Proceedings of the 42nd Conference on Education and Research in Computer Aided Architectural Design in Europe (eCAADe 2024), Nicosia, 11-13 September 2024, Volume 2, pp. 547–556
summary	This study introduces a novel data-driven approach to quantitatively characterize and measure framing and reframing (F-RF) behaviors during design problem-solving. F-RF are cognitive processes which shape problem understanding and solution development in design. Quantitative measurement methods for F-RF remain largely unexplored. The proposed approach utilizes protocol analysis combined with Natural Language Processing (NLP) algorithms to track the occurrences and re-occurrences of design concepts expressed verbally while designing. Specifically, NLP algorithms are employed to identify F-RF, enabling the systematic tracking of F-RFs and their corresponding semantic values. By calculating the semantic value of concepts and frames, the approach enables determining how a concept and a frame differed from the previous occurrences. A case study of an architect and a student demonstrates this data-driven approach. The proposed methodology holds potential for the development of systems capable of providing real-time feedback to students and professional designers, supporting and enhancing their framing skills during the design process.
keywords	Data-driven approach, Natural Language Processing (NLP), Design concept, Design problem-solving, Framing and reframing
series	eCAADe
email
last changed	2024/11/17 22:05

_id	ecaade2024_395
id	ecaade2024_395
authors	Efthimiou, Eftihis; Vaitsos, Alexandros
year	2024
title	Kaleidoscopic Pragmatism: Generative illumination of an historical Athens office building
doi	https://doi.org/10.52842/conf.ecaade.2024.2.271
source	Kontovourkis, O, Phocas, MC and Wurzer, G (eds.), Data-Driven Intelligence - Proceedings of the 42nd Conference on Education and Research in Computer Aided Architectural Design in Europe (eCAADe 2024), Nicosia, 11-13 September 2024, Volume 2, pp. 271–280
summary	This paper explores the dynamic interplay between computational design, digital simulation and physical modeling in the creation of an innovative architectural intervention within a historical office building in Athens. Spanning several decades of construction (1930s-1980s), the building's illumination waned amidst the encroachment of contemporary structures. Our solution involves a meticulously designed 2.60*2.60m light well, carving through the building's core and housing a nine-storey generative sculpture. This colossal kaleidoscopic prism, working in tandem with a heliostat, serves both as a striking aesthetic statement and an efficient lighting apparatus, ensuring stable natural lighting conditions in the surrounding office space. During the form-finding phase, rigorous computational processes guided the placement of reflective and refractive triangles within the sculpture, achieving specific degrees of porosity and performativity. Simultaneously addressing a demanding architectural program while enhancing spatial qualities, the apparatus showcases its multifaceted nature. Physical modeling played a pivotal role, with a precise model constructed from true-to-scale materials. Employing photometry, we validated lighting performance, bridging the gap between computational simulations and real-world applicability. The confluence of computational design and physical computation not only shapes the aesthetic and functional aspects of the generative sculpture but also fortifies the credibility of its lighting performance. As the work undergoes construction in Athens, this paper offers a comprehensive exploration of the innovative synergy between computational and physical methodologies, providing valuable insights into the seamless integration of cutting-edge technologies with traditional architectural practices. Within the paper, we present the intricacies of the computational apparatus, showcasing conceptual and user-friendly implementations for accessibility by non-specialist end users. We delve into digital simulation steps, integrating their findings and cross-referencing them with their physical computation counterparts, offering a holistic understanding of the project, from the standpoint of the computational designer.
keywords	generative design, performance based design, light simulation, physical modeling
series	eCAADe
email
last changed	2024/11/17 22:05

_id	caadria2024_389
id	caadria2024_389
authors	Fang, Iuan-Kai and Shih, Shen-Guan
year	2024
title	The Intersection of Technology and Architecture: Smartphone Photography in Carbon Analysis
doi	https://doi.org/10.52842/conf.caadria.2024.1.189
source	Nicole Gardner, Christiane M. Herr, Likai Wang, Hirano Toshiki, Sumbul Ahmad Khan (eds.), ACCELERATED DESIGN - Proceedings of the 29th CAADRIA Conference, Singapore, 20-26 April 2024, Volume 1, pp. 189–198
summary	Our research introduces an innovative methodology that employs smartphone imaging for measuring dimensions and utilizes deep learning to estimate carbon emissions associated with facade materials. The dimensions of various components of building exteriors are obtained through smartphone imaging, and a network model on a cloud server automatically segments these components in the images, calculating their respective areas. By combining user-input material specifications such as thickness and density? with standard values of material carbon coefficients, estimations for each component's material carbon footprint are derived. This approach offers the advantage of individual estimations for diverse materials, aiding in the design of low-carbon facades. Additionally, it features a user-friendly interface enabling swift carbon estimation through portable devices. The method provides a convenient and efficient means for assessing carbon emissions in building facades, contributing to sustainable efforts and informed material selections for a greener future.
keywords	Carbon Emission, Façade, Part Segmentation, Smartphone.
series	CAADRIA
email
last changed	2024/11/17 22:05

_id	ecaade2024_230
id	ecaade2024_230
authors	Fekar, Hugo; Novák, Jan; Míèa, Jakub; Žigmundová, Viktória; Suleimanova, Diana; Tsikoliya, Shota; Vasko, Imrich
year	2024
title	Fabrication with Residual Wood through Scanning Optimization and Robotic Milling
doi	https://doi.org/10.52842/conf.ecaade.2024.1.025
source	Kontovourkis, O, Phocas, MC and Wurzer, G (eds.), Data-Driven Intelligence - Proceedings of the 42nd Conference on Education and Research in Computer Aided Architectural Design in Europe (eCAADe 2024), Nicosia, 11-13 September 2024, Volume 1, pp. 25–34
summary	The project deals with the use of residual wood of tree stumps and roots through scanning, optimization and robotic milling. Wood logging residue makes up to 50 percent of the trees harvested biomass. (Hakkila and Parikka 2002). Among prevailing strategies is leaving residue on site, and recovering residue for bioenergy. (Perlack and others 2005). The project explores the third strategy, using parts of the logging residue for fabrication, which may reduce the overall amount of wood logging volume. Furthermore approach aims for applying residue in its natural form and taking advantage of specific local characteristics of wood (Desch and Dinwoodie 1996). The project applies the strategy on working with stump and roots of an oak tree. Due to considerations of scale, available milling technics and available resources, chosen goal of the approach is to create a functioning chair prototype. Among the problems of the approach is the complex shape of the residue, uneven quality of wood, varying humidity and contamination with soil. After cleaning and drying, the stump is scanned and a 3D model is created. The 3D model od a stump is confronted with a 3D modelled limits of the goal typology (height, width, length, sitting surface area and overal volume of a chair) and topological optimization algorithm is used to iteratively reach the desired geometry. Unlike in established topological optimization proces, which aims for a minimal volume, the project attempts to achieve required qualities with removing minimal amount of wood. Due to geometric complexity of both stump and goal object, milling with an 6axis industrial robotic arm and a rotary table was chosen as a fabrication method. The object was clamped to the board (then connected to a rotary table) in order to provide precise location and orientation in 3D space. The milling of the object was divided in two parts, with the seating area milled in higher detail. Overall process of working with a residual wood that has potential to be both effective and present aesthetic quality based on individual characteristics of wood. Further development can integrate a generative tool which would streamline the design and fabrication proces further.
keywords	Robotic arm milling, Scanning, Residual wood
series	eCAADe
email
last changed	2024/11/17 22:05

_id	caadria2024_331
id	caadria2024_331
authors	Gao, Naixiang
year	2024
title	Interactive Mediatised Urbanism: Shaping High Emotional Value Food Consumption Spaces With Human Data on Social Media
doi	https://doi.org/10.52842/conf.caadria.2024.2.415
source	Nicole Gardner, Christiane M. Herr, Likai Wang, Hirano Toshiki, Sumbul Ahmad Khan (eds.), ACCELERATED DESIGN - Proceedings of the 29th CAADRIA Conference, Singapore, 20-26 April 2024, Volume 2, pp. 415–424
summary	Research in the field of shaping urban spaces with emotional values on social media is still lacking. This paper attempts to shape an urban food consumption space with high emotional value through digital interactive media consisting of machine learning and other algorithmic processing of the emotional values contained in images of urban food consumption venues posted by humans on social media for the food consumption process. Images on social media that contain geographic information on the food market with research conditions are used as the data source, which is filtered by algorithms for the emotional value data. The filtered images are put into a machine learning model for training, resulting in a series of spatial images containing high emotional values, which are used as learning objects for the pictures entered by the user community at the input side of the system. The depth information in the output objects is read to transform them into spatial models, and the information from these virtual spatial models that have been learned for higher emotional values is the output of user interactions and the basis for improving the food consumption space.
keywords	social media database, human-urban interactive media, computer visualization, image-to-model, food consumption space, emotional values, machine learning, AR interface
series	CAADRIA
email
last changed	2024/11/17 22:05

For more results click below: