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 576

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

_id	ecaade2023_44
id	ecaade2023_44
authors	Mayrhofer-Hufnagl, Ingrid and Ennemoser, Benjamin
year	2023
title	From Linear to Manifold Interpolation: Exemplifying the paradigm shift through interpolation
doi	https://doi.org/10.52842/conf.ecaade.2023.2.419
source	Dokonal, W, Hirschberg, U and Wurzer, G (eds.), Digital Design Reconsidered - Proceedings of the 41st Conference on Education and Research in Computer Aided Architectural Design in Europe (eCAADe 2023) - Volume 2, Graz, 20-22 September 2023, pp. 419–429
summary	The advent of artificial intelligence, specifically neural networks, has marked a significant turning point in the field of computation. During such transformative times, we are often faced with a dearth of appropriate vocabulary, which forces us to rely on existing terms, regardless of their inadequacy. This paper argues that the term “interpolation,” typically used in deep learning (DL), is a prime example of this phenomenon. It is not uncommon for beginners to misunderstand its meaning, as DL pioneer Francois Chollet (2017) has noted. This misreading is especially true in the discipline of architecture, and this study aims to demonstrate how the meaning of “interpolation” has evolved in the second digital turn. We begin by illustrating, using 2D data, the difference between linear interpolation in the context of topological figures and its use in DL algorithms. We then demonstrate how 3DGANs can be employed to interpolate across different topologies in complex 3D space, highlighting the distinction between linear and manifold interpolation. In both 2D and 3D examples, our results indicate that the process does not involve continuous morphing but instead resembles the piecing together of a jigsaw puzzle to form many parts of a larger ambient space. Our study reveals how previous architectural research on DL has employed the term “interpolation” without clarifying the crucial differences from its use in the first digital turn. We demonstrate the new possibilities that manifold interpolation offers for architecture, which extend well beyond parametric variations of the same topology.
keywords	Interpolation, 3D Generative Adversarial Networks, Deep Learning, Hybrid Space
series	eCAADe
email
last changed	2023/12/10 10:49

_id	acadia17_552
id	acadia17_552
authors	Sjoberg, Christian; Beorkrem, Christopher; Ellinger, Jefferson
year	2017
title	Emergent Syntax: Machine Learning for the Curation of Design Solution Space
doi	https://doi.org/10.52842/conf.acadia.2017.552
source	ACADIA 2017: DISCIPLINES & DISRUPTION [Proceedings of the 37th Annual Conference of the Association for Computer Aided Design in Architecture (ACADIA) ISBN 978-0-692-96506-1] Cambridge, MA 2-4 November, 2017), pp. 552- 561
summary	The expanding role of computational models in the process of design is producing exponential growth in parameter spaces. As designers, we must create and implement new methods for searching these parameter spaces, considering not only quantitative optimization metrics but also qualitative features. This paper proposes a methodology that leverages the pattern modeling properties of artificial neural networks to capture designers' inexplicit selection criteria and create user-selection-based fitness functions for a genetic solver. Through emulation of learned selection patterns, fitness functions based on trained networks provide a method for qualitative evaluation of designs in the context of a given population. The application of genetic solvers for the generation of new populations based on the trained network selections creates emergent high-density clusters in the parameter space, allowing for the identification of solutions that satisfy the designer’s inexplicit criteria. The results of an initial user study show that even with small numbers of training objects, a search tool with this configuration can begin to emulate the design criteria of the user who trained it.
keywords	design methods; information processing; AI; machine learning; generative system
series	ACADIA
email
last changed	2022/06/07 07:56

_id	cf2019_009
id	cf2019_009
authors	Veloso, Pedro; Jinmo Rhee and Ramesh Krishnamurti
year	2019
title	Multi-agent space planning: a literature review (2008-2017)
source	Ji-Hyun Lee (Eds.) "Hello, Culture!" [18th International Conference, CAAD Futures 2019, Proceedings / ISBN 978-89-89453-05-5] Daejeon, Korea, pp. 52-74
summary	In this paper we review the research on multi-agent space planning (MASP) during the period of 2008-2017. By MASP, we refer to space planning (SP) methods based on online mobile agents that map local perceptions to actions in the environment, generating spatial representation. We group two precedents and sixteen recent MASP prototypes into three categories: (1) agents as moving spatial units, (2) agents that occupy a space, and (3) agents that partition a space. In order to compare the prototypes, we identify the occurrence of features in terms of representation, objectives, and control procedures. Upon analysis of occurrences and correlations of features in the types, we present gaps and challenges for future MASP research. We point to the limits of current systems to solve spatial conflicts and to incorporate architectural knowledge. Finally, we suggest that behavioral learning offers a promising path for robust and autonomous MASP systems in the architectural domain.
keywords	Space planning; Agent-based modeling; Multi-agent systems; Generative systems
series	CAAD Futures
email
last changed	2019/07/29 14:08

_id	ijac201715101
id	ijac201715101
authors	Bieg, Kory and Clay Odom
year	2017
title	Lumifoil and Tschumi: Virtual projections and architectural interventions
source	International Journal of Architectural Computing vol. 15 - no. 1, 6-17
summary	This article introduces the theoretical and technical framework for the design of a temporary rooftop canopy on the red generator—one of the buildings designed by Bernard Tschumi for the Florida International University School of Architecture. The project, Lumifoil, was designed using both top-down and bottom-up computational techniques, including surface modeling via projected geometries and scripted cellular subdivisions and assemblies. Lumifoil attempts to synthesize these two often-conflicting design approaches into a generative design process which leverages context, form, surface, and structure as affective and effective actors. Lumifoil is the result of a design methodology which is both active and reactive to existing conditions of the site and new opportunities afforded by the program. It is contextual in its top-down relationship to Tschumi’s existing building and theory, generative in how details emerge bottom-up through scripts which lack any reference to site, and emergent in the resulting synthetic processes and effects which are produced. Through this methodological development, the project both tracks and responds to popular architectural theory and design from the mid-1990s to today. The theoretical underpinnings of the project build upon the idea that the actual (the real-life physical manifestation of matter) and the virtual (the potential for an object to be) are two constantly shifting paradigms in which design processes can intervene to help develop an architectural solution from a range of possibilities. The technical aspect of the project includes the collaborative workflow between the architecture offices of OTA+ and studio MODO with Arup Engineers to resolve structural issues using parametric modeling tools and structural analysis software. The final project is entirely parametric and fabrication is completely automated.
keywords	Tschumi, Parametric, Installation, Generative, Projection
series	other
type	normal paper
email
last changed	2019/08/02 08:16

_id	caadria2017_095
id	caadria2017_095
authors	Lee, Hyo Jung and Lee, Hyunsoo
year	2017
title	Automatic 3D Modeling of Korean Traditional Architecture - Applying Parametric Design
doi	https://doi.org/10.52842/conf.caadria.2017.231
source	P. Janssen, P. Loh, A. Raonic, M. A. Schnabel (eds.), Protocols, Flows, and Glitches - Proceedings of the 22nd CAADRIA Conference, Xi'an Jiaotong-Liverpool University, Suzhou, China, 5-8 April 2017, pp. 231-240
summary	Korean traditional structure is constructed as prefabrication jointed in largely characterized by its unique components under the specific rules of assembly and proportion. This point is a double-edges sword. Because, while various shapes and sizes of components based upon an objected-oriented form appear the potential possibility of producing changeable prototypes to build up, these various characters of components and several jointed methods has made difficulties to handle. Accordingly, an automatic 3D modeling algorithm is focused on the methodology of changeable prototypes of Korean Traditional architecture keeping traditional jointed methods with setting various characters of components
keywords	Korean traditional structure; Parametric design ; Generative three dimensional modeling ; Hanok.
series	CAADRIA
email
last changed	2022/06/07 07:51

_id	caadria2017_129
id	caadria2017_129
authors	Patt, Trevor Ryan
year	2017
title	Toward Temporal and Punctual Urban Redevelopment in Dynamic, Informal Contexts - An Adaptive Masterplan Driven by Architectural Interventions Using Multiagent Modeling
doi	https://doi.org/10.52842/conf.caadria.2017.221
source	P. Janssen, P. Loh, A. Raonic, M. A. Schnabel (eds.), Protocols, Flows, and Glitches - Proceedings of the 22nd CAADRIA Conference, Xi'an Jiaotong-Liverpool University, Suzhou, China, 5-8 April 2017, pp. 221-230
summary	This paper presents design research speculating on new planning approaches for informal urban sites that enables coordinated planning to operate within the realm of spontaneous, bottom-up redevelopment. In opposition to the /tabula rasa/ Modernist development, this project reacts to the dynamic metabolism of the village and engages with the rapid turnover of the built environment of the village as a mechanism through which to implement incremental redevelopment. A radical reorientation of the object of masterplanning, this replaces the singular image or document as the guiding authority with a collection of opportunistic adaptations, temporal sequences, and localized procedures. Enabling this approach is a computational approach that analyzes the morphology of the public space network to identify opportunities to address issues in the composition of the village. A multiagent system driven by weighted random walks through the circulation network conducts local analyses of the urban fabric while changes are made and proposes potential modifications to discrete areas. The model simulates the potential for such a planning tool to be used over a long time span and updated with empirically gathered data, having the benefit of flexibility and resilience in the face of the changing and unregulated conditions in the context of informal urbanism.
keywords	generative design; responsive masterplanning; informal urbanism; network analysis; agent-based modeling
series	CAADRIA
email
last changed	2022/06/07 07:59

_id	acadia17_512
id	acadia17_512
authors	Rossi, Andrea; Tessmann, Oliver
year	2017
title	Collaborative Assembly of Digital Materials
doi	https://doi.org/10.52842/conf.acadia.2017.512
source	ACADIA 2017: DISCIPLINES & DISRUPTION [Proceedings of the 37th Annual Conference of the Association for Computer Aided Design in Architecture (ACADIA) ISBN 978-0-692-96506-1] Cambridge, MA 2-4 November, 2017), pp. 512- 521
summary	Current developments in design-to-production workflows aim to allow architects to quickly prototype designs that result from advanced design processes while also embedding the constraints imposed by selected fabrication equipment. However, the enduring physical separation between design space and fabrication space, together with a continuous approach to both design, via NURBs modeling software, and fabrication, through irreversible material processing methods, limit the possibilities to extend the advantages of a “digital” approach (Ward 2010), such as full editability and reversibility, to physical realizations. In response to such issues, this paper proposes a processto allow the concurrent design and fabrication of discrete structures in a collaborative process between human designer and a 6-axis robotic arm. This requires the development of design and materialization procedures for discrete aggregations, including the modeling of assembly constraints, as well as the establishment of a communication platform between human and machine actors. This intends to offer methods to increase the accessibility of discrete design methodologies, as well as to hint at possibilities for overcoming the division between design and manufacturing (Carpo 2011; Bard et al. 2014), thus allowing intuitive design decisions to be integrated directly within assembly processes (Johns 2014).
keywords	material and construction; construction/robotics; smart assembly/construction; generative system
series	ACADIA
email
last changed	2022/06/07 07:56

_id	ecaade2017_021
id	ecaade2017_021
authors	Agirbas, Asli
year	2017
title	The Use of Simulation for Creating Folding Structures - A Teaching Model
doi	https://doi.org/10.52842/conf.ecaade.2017.1.325
source	Fioravanti, A, Cursi, S, Elahmar, S, Gargaro, S, Loffreda, G, Novembri, G, Trento, A (eds.), ShoCK! - Sharing Computational Knowledge! - Proceedings of the 35th eCAADe Conference - Volume 1, Sapienza University of Rome, Rome, Italy, 20-22 September 2017, pp. 325-332
summary	In architectural education, the demand for creating forms with a non-Euclidean geometry, which can only be achieved by using the computer-aided design tools, is increasing. The teaching of this subject is a great challenge for both students and instructors, because of the intensive nature of architecture undergraduate programs. Therefore, for the creation of those forms with a non-Euclidean geometry, experimental work was carried out in an elective course based on the learning visual programming language. The creation of folding structures with form-finding by simulation was chosen as the subject of the design production which would be done as part of the content of the course. In this particular course, it was intended that all stages should be experienced, from the modeling in the virtual environment to the digital fabrication. Hence, in their early years of architectural education, the students were able to learn versatile thinking by experiencing, simultaneously, the use of simulation in the environment of visual programming language, the forming space by using folding structures, the material-based thinking and the creation of their designs suitable to the digital fabrication.
keywords	Folding Structures; CAAD; Simulation; Form-finding; Architectural Education
series	eCAADe
email
last changed	2022/06/07 07:54

_id	ecaade2017_037
id	ecaade2017_037
authors	Hassan Khalil, Mohamed
year	2017
title	Learning by Merging 3D Modeling for CAAD with the Interactive Applications - Bearing walls, Vaults, Domes as Case study
doi	https://doi.org/10.52842/conf.ecaade.2017.1.353
source	Fioravanti, A, Cursi, S, Elahmar, S, Gargaro, S, Loffreda, G, Novembri, G, Trento, A (eds.), ShoCK! - Sharing Computational Knowledge! - Proceedings of the 35th eCAADe Conference - Volume 1, Sapienza University of Rome, Rome, Italy, 20-22 September 2017, pp. 353-362
summary	The development and the innovation of tools, techniques and digital applications represent a challenge for those who are in charge of architectural education to keep up with this development. This is because these techniques provide potentials that are not available in the traditional method of teaching. This raises an important question: can these tools and techniques help to achieve the targeted outcomes of education? This research paper discusses how to integrate both digital 3D models, of CAAD, and interactive applications for the development of architectural education curriculum. To test this, a case study has been conducted on the subject of building construction, for the second year at the faculty of engineering, specifically, the bearing walls construction system. In addition, this study has been divided into three parts. Through the first part, the scientific content of the curriculum, which tackles the bearing walls, has been prepared. The second part shows how to convert the scientific content into an interactive content in which the students learn through the experiment and the simulation of the traditional construction methods as the students a acquire construction skills and the ability to imagine different structural complexities. The third part includes the creation of both the application and the software containing the interactive curriculum. Workshop for the students has been held as a case study to test the effectiveness of this development and to recognize the pros and cons. The results confirmed the importance of integrating this applications into architectural education.
keywords	CAAD; 3D modeling ; Building Construction; Interactive applications; Bearing walls systems
series	eCAADe
email
last changed	2022/06/07 07:49

_id	acadia20_382
id	acadia20_382
authors	Hosmer, Tyson; Tigas, Panagiotis; Reeves, David; He, Ziming
year	2020
title	Spatial Assembly with Self-Play Reinforcement Learning
doi	https://doi.org/10.52842/conf.acadia.2020.1.382
source	ACADIA 2020: Distributed Proximities / Volume I: Technical Papers [Proceedings of the 40th Annual Conference of the Association of Computer Aided Design in Architecture (ACADIA) ISBN 978-0-578-95213-0]. Online and Global. 24-30 October 2020. edited by B. Slocum, V. Ago, S. Doyle, A. Marcus, M. Yablonina, and M. del Campo. 382-393.
summary	We present a framework to generate intelligent spatial assemblies from sets of digitally encoded spatial parts designed by the architect with embedded principles of prefabrication, assembly awareness, and reconfigurability. The methodology includes a bespoke constraint-solving algorithm for autonomously assembling 3D geometries into larger spatial compositions for the built environment. A series of graph-based analysis methods are applied to each assembly to extract performance metrics related to architectural space-making goals, including structural stability, material density, spatial segmentation, connectivity, and spatial distribution. Together with the constraint-based assembly algorithm and analysis methods, we have integrated a novel application of deep reinforcement (RL) learning for training the models to improve at matching the multiperformance goals established by the user through self-play. RL is applied to improve the selection and sequencing of parts while considering local and global objectives. The user’s design intent is embedded through the design of partial units of 3D space with embedded fabrication principles and their relational constraints over how they connect to each other and the quantifiable goals to drive the distribution of effective features. The methodology has been developed over three years through three case study projects called ArchiGo (2017–2018), NoMAS (2018–2019), and IRSILA (2019-2020). Each demonstrates the potential for buildings with reconfigurable and adaptive life cycles.
series	ACADIA
type	paper
email
last changed	2023/10/22 12:06

_id	caadria2023_362
id	caadria2023_362
authors	Luo, Jiaxiang, Mastrokalou, Efthymia, Aldabous, Rahaf, Aldaboos, Sarah and Lopez Rodriguez, Alvaro
year	2023
title	Fabrication of Complex Clay Structures Through an Augmented Reality Assisted Platform
doi	https://doi.org/10.52842/conf.caadria.2023.1.413
source	Immanuel Koh, Dagmar Reinhardt, Mohammed Makki, Mona Khakhar, Nic Bao (eds.), HUMAN-CENTRIC - Proceedings of the 28th CAADRIA Conference, Ahmedabad, 18-24 March 2023, pp. 413–422
summary	The relationship between clay manufacturing and architectural design has a long trajectory that has been explored since the early 2000s. From a 3D printing or assembly perspective, using clay in combination with automated processes in architecture to achieve computational design solutions is well established. (Yuan, Leach & Menges, 2018). Craft-based clay art, however, still lacks effective computational design integration. With the improvement of Augmented Reality (AR) technologies (Driscoll et al., 2017) and the appearance of digital platforms, new opportunities to integrate clay manufacturing and computational design have emerged. The concept of digitally transferring crafting skills, using holographic guidance and machine learning, could make clay crafting accessible to more workers while creating the potential to share and exchange digital designs via an open-source manufacturing platform. In this context, this research project explores the potential of integrating computational design and clay crafting using AR. Moreover, it introduces a platform that enables AR guidance and the digital transfer of fabrication skills, allowing even amateur users with no prior making experience to produce complex clay components.
keywords	Computer vision, Distributed manufacturing, Augmented craftsmanship, Augmented reality, Real-time modification, Hololens
series	CAADRIA
email
last changed	2023/06/15 23:14

_id	acadia17_426
id	acadia17_426
authors	Moorman, Andrew
year	2017
title	Pattern Making and Learning: Non-Routine Practices in Generative Design
doi	https://doi.org/10.52842/conf.acadia.2017.426
source	ACADIA 2017: DISCIPLINES & DISRUPTION [Proceedings of the 37th Annual Conference of the Association for Computer Aided Design in Architecture (ACADIA) ISBN 978-0-692-96506-1] Cambridge, MA 2-4 November, 2017), pp. 426- 435
summary	We now witness an upsurge in mainstream generative design tools fortified by simulation that speed up the concealed linear synthesis of optimized design alternatives. In pursuit of optimality, these tools saturate local machines or cloud servers with analysis and design iteration data, only to discard it once the procedure has concluded. Largely absent, however, are tools for an active, adaptive relationship with design exploration and the reuse of corresponding design data and metadata. In Pattern Making and Pattern Learning, we propose that these characteristics are mutually beneficial. This paper presents a series of revisions to the optimization framework for routine design synthesis that examine a potential symbiosis between the production of large datasets (big data) and non-routine practices of making in design. Our engagement with iterative design exercises is twofold: as a supply of computer-generated design information to foster user intuition and explore the design space on non-objective terms, and as a supply of human-generated design information to learn artifacts of user preference in the interest of design software personalization. These concepts are applied to the generation of functionally graded patterning in chair design, combining methods of physical production with programmable sheet material behavior through a custom interactive synthesis framework.
keywords	design methods; information processing; ai & machine learning; simulation & optimization; generative system
series	ACADIA
email
last changed	2022/06/07 07:58

_id	ecaade2017_232
id	ecaade2017_232
authors	Ostrowska-Wawryniuk, Karolina, Markusiewicz, Jacek and S³yk, Jan
year	2017
title	Descriptive Geometry 2.0 - Define vs. design
doi	https://doi.org/10.52842/conf.ecaade.2017.2.425
source	Fioravanti, A, Cursi, S, Elahmar, S, Gargaro, S, Loffreda, G, Novembri, G, Trento, A (eds.), ShoCK! - Sharing Computational Knowledge! - Proceedings of the 35th eCAADe Conference - Volume 2, Sapienza University of Rome, Rome, Italy, 20-22 September 2017, pp. 425-430
summary	The article presents the 'Digital Geometry Techniques' course taught at the second year of the undergraduate course at the Faculty of Architecture in the Warsaw University of Technology - WUT. The course introduces mathematical theory and generative modeling in order to prepare the students to consciously plan their creative process and to choose the set of tools according to an initial analysis of modeling constraints. The students gain knowledge on advanced CAAD techniques through learning functions of a particular program, and also by tackling geometry-related problems derived from real-world architectural projects. They are able to develop individual solutions using adequate techniques. We present three different students' semester works as examples to reflect on the significance of mathematics and algorithmization in the process of problem solving and form creation in architecture and urban design.
keywords	project based learning; generative design; architectural curriculum; conceptual thinking; geometry; programming
series	eCAADe
email
last changed	2022/06/07 08:00

_id	acadia17_474
id	acadia17_474
authors	Peng, Wenzhe; Zhang, Fan; Nagakura, Takehiko
year	2017
title	Machines’ Perception of Space: Employing 3D Isovist Methods and a Convolutional Neural Network in Architectural Space Classification
doi	https://doi.org/10.52842/conf.acadia.2017.474
source	ACADIA 2017: DISCIPLINES & DISRUPTION [Proceedings of the 37th Annual Conference of the Association for Computer Aided Design in Architecture (ACADIA) ISBN 978-0-692-96506-1] Cambridge, MA 2-4 November, 2017), pp. 474- 481
summary	Simple and common architectural elements can be combined to create complex spaces. Different spatial compositions of elements define different spatial boundaries, and each produces a unique local spatial experience to observers inside the space. Therefore an architectural style brings about a distinct spatial experience. While multiple representation methods are practiced in the field of architecture, there lacks a compelling way to capture and identify spatial experiences. Describing an observer’s spatial experiences quantitatively and efficiently is a challenge. In this paper, we propose a method that employs 3D isovist methods and a convolutional neural network (CNN) to achieve recognition of local spatial compositions. The case studies conducted validate that this methodology works well in capturing and identifying local spatial conditions, illustrates the pattern and frequency of their appearance in designs, and indicates peculiar spatial experiences embedded in an architectural style. The case study used small designs by Mies van der Rohe and Aldo van Eyck. The contribution of this paper is threefold. First, it introduces a sampling method based on 3D Isovist that generates a 2D image that can be used to represent a 3D space from a specific observation point. Second, it employs a CNN model to extract features from the sampled images, then classifies their corresponding space. Third, it demonstrates a few case studies where this space classification method is applied to different architectural styles.
keywords	design methods; information processing; AI; machine learning; computer vision; representation
series	ACADIA
email
last changed	2022/06/07 08:00

_id	ecaade2017_269
id	ecaade2017_269
authors	Rahmani Asl, Mohammad, Das, Subhajit, Tsai, Barry, Molloy, Ian and Hauck, Anthony
year	2017
title	Energy Model Machine (EMM) - Instant Building Energy Prediction using Machine Learning
doi	https://doi.org/10.52842/conf.ecaade.2017.2.277
source	Fioravanti, A, Cursi, S, Elahmar, S, Gargaro, S, Loffreda, G, Novembri, G, Trento, A (eds.), ShoCK! - Sharing Computational Knowledge! - Proceedings of the 35th eCAADe Conference - Volume 2, Sapienza University of Rome, Rome, Italy, 20-22 September 2017, pp. 277-286
summary	In the process of building design, energy performance is often simulated using physical principles of thermodynamics and energy behaviour using elaborate simulation tools. However, energy simulation is computationally expensive and time consuming process. These drawbacks limit opportunities for design space exploration and prevent interactive design which results in environmentally inefficient buildings. In this paper we propose Energy Model Machine (EMM) as a general and flexible approximation model for instant energy performance prediction using machine learning (ML) algorithms to facilitate design space exploration in building design process. EMM can easily be added to design tools and provide instant feedback for real-time design iterations. To demonstrate its applicability, EMM is used to estimate energy performance of a medium size office building during the design space exploration in widely used parametrically design tool as a case study. The results of this study support the feasibility of using machine learning approaches to estimate energy performance for design exploration and optimization workflows to achieve high performance buildings.
keywords	Machine Learning; Artificial Neural Networks; Boosted Decision Tree; Building Energy Performance; Parametric Modeling and Design; Building Performance Optimization
series	eCAADe
email
last changed	2022/06/07 08:00

_id	ecaade2023_259
id	ecaade2023_259
authors	Sonne-Frederiksen, Povl Filip, Larsen, Niels Martin and Buthke, Jan
year	2023
title	Point Cloud Segmentation for Building Reuse - Construction of digital twins in early phase building reuse projects
doi	https://doi.org/10.52842/conf.ecaade.2023.2.327
source	Dokonal, W, Hirschberg, U and Wurzer, G (eds.), Digital Design Reconsidered - Proceedings of the 41st Conference on Education and Research in Computer Aided Architectural Design in Europe (eCAADe 2023) - Volume 2, Graz, 20-22 September 2023, pp. 327–336
summary	Point cloud processing has come a long way in the past years. Advances in computer vision (CV) and machine learning (ML) have enabled its automated recognition and processing. However, few of those developments have made it through to the Architecture, Engineering and Construction (AEC) industry. Here, optimizing those workflows can reduce time spent on early-phase projects, which otherwise could be spent on developing innovative design solutions. Simplifying the processing of building point cloud scans makes it more accessible and therefore, usable for design, planning and decision-making. Furthermore, automated processing can also ensure that point clouds are processed consistently and accurately, reducing the potential for human error. This work is part of a larger effort to optimize early-phase design processes to promote the reuse of vacant buildings. It focuses on technical solutions to automate the reconstruction of point clouds into a digital twin as a simplified solid 3D element model. In this paper, various ML approaches, among others KPConv Thomas et al. (2019), ShapeConv Cao et al. (2021) and Mask-RCNN He et al. (2017), are compared in their ability to apply semantic as well as instance segmentation to point clouds. Further it relies on the S3DIS Armeni et al. (2017), NYU v2 Silberman et al. (2012) and Matterport Ramakrishnan et al. (2021) data sets for training. Here, the authors aim to establish a workflow that reduces the effort for users to process their point clouds and obtain object-based models. The findings of this research show that although pure point cloud-based ML models enable a greater degree of flexibility, they incur a high computational cost. We found, that using RGB-D images for classifications and segmentation simplifies the complexity of the ML model but leads to additional requirements for the data set. These can be mitigated in the initial process of capturing the building or by extracting the depth data from the point cloud.
keywords	Point Clouds, Machine Learning, Segmentation, Reuse, Digital Twins
series	eCAADe
email
last changed	2023/12/10 10:49

_id	acadia19_392
id	acadia19_392
authors	Steinfeld, Kyle
year	2019
title	GAN Loci
doi	https://doi.org/10.52842/conf.acadia.2019.392
source	ACADIA 19:UBIQUITY AND AUTONOMY [Proceedings of the 39th Annual Conference of the Association for Computer Aided Design in Architecture (ACADIA) ISBN 978-0-578-59179-7] (The University of Texas at Austin School of Architecture, Austin, Texas 21-26 October, 2019) pp. 392-403
summary	This project applies techniques in machine learning, specifically generative adversarial networks (or GANs), to produce synthetic images intended to capture the predominant visual properties of urban places. We propose that imaging cities in this manner represents the first computational approach to documenting the Genius Loci of a city (Norberg-Schulz, 1980), which is understood to include those forms, textures, colors, and qualities of light that exemplify a particular urban location and that set it apart from similar places. Presented here are methods for the collection of urban image data, for the necessary processing and formatting of this data, and for the training of two known computational statistical models (StyleGAN (Karras et al., 2018) and Pix2Pix (Isola et al., 2016)) that identify visual patterns distinct to a given site and that reproduce these patterns to generate new images. These methods have been applied to image nine distinct urban contexts across six cities in the US and Europe, the results of which are presented here. While the product of this work is not a tool for the design of cities or building forms, but rather a method for the synthetic imaging of existing places, we nevertheless seek to situate the work in terms of computer-assisted design (CAD). In this regard, the project is demonstrative of a new approach to CAD tools. In contrast with existing tools that seek to capture the explicit intention of their user (Aish, Glynn, Sheil 2017), in applying computational statistical methods to the production of images that speak to the implicit qualities that constitute a place, this project demonstrates the unique advantages offered by such methods in capturing and expressing the tacit.
series	ACADIA
type	normal paper
email
last changed	2022/06/07 07:56

_id	ecaade2017_009
id	ecaade2017_009
authors	Takizawa, Atsushi and Furuta, Airi
year	2017
title	3D Spatial Analysis Method with First-Person Viewpoint by Deep Convolutional Neural Network with Omnidirectional RGB and Depth Images
doi	https://doi.org/10.52842/conf.ecaade.2017.2.693
source	Fioravanti, A, Cursi, S, Elahmar, S, Gargaro, S, Loffreda, G, Novembri, G, Trento, A (eds.), ShoCK! - Sharing Computational Knowledge! - Proceedings of the 35th eCAADe Conference - Volume 2, Sapienza University of Rome, Rome, Italy, 20-22 September 2017, pp. 693-702
summary	The fields of architecture and urban planning widely apply spatial analysis based on images. However, many features can influence the spatial conditions, not all of which can be explicitly defined. In this research, we propose a new deep learning framework for extracting spatial features without explicitly specifying them and use these features for spatial analysis and prediction. As a first step, we establish a deep convolution neural network (DCNN) learning problem with omnidirectional images that include depth images as well as ordinary RGB images. We then use these images as explanatory variables in a game engine to predict a subjects' preference regarding a virtual urban space. DCNNs learn the relationship between the evaluation result and the omnidirectional camera images and we confirm the prediction accuracy of the verification data.
keywords	Space evaluation; deep convolutional neural network; omnidirectional image; depth image; Unity; virtual reality
series	eCAADe
email
last changed	2022/06/07 07:56

_id	ecaade2017_183
id	ecaade2017_183
authors	Wendell, Augustus and Altin, Ersin
year	2017
title	Learning Space - Incorporating spatial simulations in design history coursework
doi	https://doi.org/10.52842/conf.ecaade.2017.1.261
source	Fioravanti, A, Cursi, S, Elahmar, S, Gargaro, S, Loffreda, G, Novembri, G, Trento, A (eds.), ShoCK! - Sharing Computational Knowledge! - Proceedings of the 35th eCAADe Conference - Volume 1, Sapienza University of Rome, Rome, Italy, 20-22 September 2017, pp. 261-266
summary	Art and architectural history education has long relied on photographic imagery. The geography of architectural history often demands an analog representation for the built form and photographic recordings have long been the widely adopted standard. In many cases, specific buildings have been taught for generations based on a handful of historical exposures. The impact of this precedent is an imperfect and highly privileged conception of architectural forms. Students learn only of a particular viewpoint of any given building, rather than understanding the building as a whole. Augmenting the tradition of select and static imagery in the classroom with new technologies can create a more comprehensive understanding of architectural precedents. This paper discusses an experiment conducted in Spring 2017 in presenting an architectural case study to a history class using a Virtual Reality 3D experience in comparison to a set of canonical photographs.
keywords	Unreal Engine; Virtual Reality; Photography; 3D; Education
series	eCAADe
email
last changed	2022/06/07 07:58

_id	acadia17_72
id	acadia17_72
authors	Alfaiate, Pedro; Caetano, In?s; Leit?o, António
year	2017
title	Luna Moth: Supporting Creativity in the Cloud
doi	https://doi.org/10.52842/conf.acadia.2017.072
source	ACADIA 2017: DISCIPLINES & DISRUPTION [Proceedings of the 37th Annual Conference of the Association for Computer Aided Design in Architecture (ACADIA) ISBN 978-0-692-96506-1] Cambridge, MA 2-4 November, 2017), pp. 72-81
summary	Algorithmic design allows architects to design using a programming-based approach. Current algorithmic design environments are based on existing computer-aided design applications or building information modeling applications, such as AutoCAD, Rhinoceros 3D, or Revit, which, due to their complexity, fail to give architects the immediate feedback they need to explore algorithmic design. In addition, they do not address the current trend of moving applications to the cloud to improve their availability. To address these problems, we propose a software architecture for an algorithmic design integrated development environment (IDE), based on web technologies, that is more interactive than competing algorithmic design IDEs. Besides providing an intuitive editing interface which facilitates programming tasks for architects, its performance can be an order of magnitude faster than current algorithmic design IDEs, thus supporting real-time feedback with more complex algorithmic design programs. Moreover, our solution also allows architects to export the generated model to their preferred computer-aided design applications. This results in an algorithmic design environment that is accessible from any computer, while offering an interactive editing environment that integrates into the architect’s workflow.
keywords	design methods; information processing; generative system; computational / artistic cultures
series	ACADIA
email
last changed	2022/06/07 07:54

For more results click below: