“…Due to the stochastic nature of these algorithms, the adjacency constraints specified by the users are not necessarily satisfied. House‐GAN++ [NHC*21] proposes compatibility as a metric to evaluate the floorplans. It is computed by calculating the graph‐edit distance between the input graph and the estimated graph from the generated floorplan.…”
Section: Results and Comparisonsmentioning
confidence: 99%
“…Due to the deterministic nature of our algorithm, G2PLAN ensures that user adjacency is maintained, thus achieving a compatibility score of 0. In contrast, on 1000 samples divided into four groups on basis of number of rooms (5,6,7,8), state‐of‐the‐art methods like House‐GAN++ achieve a compatibility score of 1.9 [NHC*21]. Figure 15 shows input graphs and the floorplans achieved using state‐of‐the‐art data‐driven methods Graph2Plan [HHT*20] and House‐GAN++, and G2PLAN.…”
Section: Results and Comparisonsmentioning
confidence: 99%
“…[ZWG*21] provide a dataset of residential plans in rural scenarios and also propose a deep learning method to simultaneously extract adjacency graphs and room functionalities from standard floorplan images. House‐GAN++[NHC*21] proposes using a conditional and a relational generative adversarial network (GAN) for automated floorplan generation, where a previously generated output floorplan becomes the next input constraint resulting in a constantly refined training process. Wamiq et al .…”
In recent times, researchers have proposed several approaches for building floorplans using parametric/generative design, shape grammars, machine learning, AI, etc. This paper aims to demonstrate a mathematical approach for the automated generation of floorplan layouts. Mathematical formulations warrant the fulfilment of all input user constraints, unlike the learning‐based methods present in the literature. Moreover, the algorithms illustrated in this paper are robust, scalable and highly efficient, generating thousands of floorplans in a few milliseconds.
We present G2PLAN, a software based on graph‐theoretic and linear optimization techniques, that generates all topologically distinct floorplans with different boundary rooms in linear time for given adjacency and dimensional constraints. G2PLAN builds on the work of GPLAN and offers solutions to a wider range of adjacency relations (one‐connected, non‐triangulated graphs) and better dimensioning customizability. It also generates a catalogue of dimensionless as well as dimensioned floorplans satisfying user requirements.
“…Due to the stochastic nature of these algorithms, the adjacency constraints specified by the users are not necessarily satisfied. House‐GAN++ [NHC*21] proposes compatibility as a metric to evaluate the floorplans. It is computed by calculating the graph‐edit distance between the input graph and the estimated graph from the generated floorplan.…”
Section: Results and Comparisonsmentioning
confidence: 99%
“…Due to the deterministic nature of our algorithm, G2PLAN ensures that user adjacency is maintained, thus achieving a compatibility score of 0. In contrast, on 1000 samples divided into four groups on basis of number of rooms (5,6,7,8), state‐of‐the‐art methods like House‐GAN++ achieve a compatibility score of 1.9 [NHC*21]. Figure 15 shows input graphs and the floorplans achieved using state‐of‐the‐art data‐driven methods Graph2Plan [HHT*20] and House‐GAN++, and G2PLAN.…”
Section: Results and Comparisonsmentioning
confidence: 99%
“…[ZWG*21] provide a dataset of residential plans in rural scenarios and also propose a deep learning method to simultaneously extract adjacency graphs and room functionalities from standard floorplan images. House‐GAN++[NHC*21] proposes using a conditional and a relational generative adversarial network (GAN) for automated floorplan generation, where a previously generated output floorplan becomes the next input constraint resulting in a constantly refined training process. Wamiq et al .…”
In recent times, researchers have proposed several approaches for building floorplans using parametric/generative design, shape grammars, machine learning, AI, etc. This paper aims to demonstrate a mathematical approach for the automated generation of floorplan layouts. Mathematical formulations warrant the fulfilment of all input user constraints, unlike the learning‐based methods present in the literature. Moreover, the algorithms illustrated in this paper are robust, scalable and highly efficient, generating thousands of floorplans in a few milliseconds.
We present G2PLAN, a software based on graph‐theoretic and linear optimization techniques, that generates all topologically distinct floorplans with different boundary rooms in linear time for given adjacency and dimensional constraints. G2PLAN builds on the work of GPLAN and offers solutions to a wider range of adjacency relations (one‐connected, non‐triangulated graphs) and better dimensioning customizability. It also generates a catalogue of dimensionless as well as dimensioned floorplans satisfying user requirements.
“…By contrast, we propose to leverage the underlying relational information between rooms to learn topological information from room attributes. The learned topological information can then facilitate indoor understanding (e.g., LayoutGMN [21]) and more home-related applications such as floorplan generation (e.g., HouseGAN [19], HouseGAN++ [20]).…”
Section: Related Work 21 Topological Relationships Of 2d/3d Floorplansmentioning
Analysis of indoor spaces requires topological information. In this paper, we propose to extract topological information from room attributes using what we call Iterative and adaptive graph Topology Learning (ITL). ITL progressively predicts multiple relations between rooms; at each iteration, it improves node embeddings, which in turn facilitates generation of a better topological graph structure. This notion of iterative improvement of node embeddings and topological graph structure is in the same spirit as [5]. However, while [5] computes the adjacency matrix based on node similarity, we learn the graph metric using a relational decoder to extract room correlations. Experiments using a new challenging indoor dataset validate our proposed method. Qualitative and quantitative evaluation for layout topology prediction and floorplan generation applications also demonstrate the effectiveness of ITL.
CCS CONCEPTS• Computing methodologies → Scene understanding; Hierarchical representations.
“…This translation is a long-standing problem, which relied on human intervention in early generative experiments (Weinzapfel & Negroponte, 1976) and on techniques based on graph embedding and triangulation (Nourian et al, 2013). Recently, this problem has been addressed with DL techniques (Nauata et al, 2020(Nauata et al, , 2021. For students to customise bubble diagrams and explore variations of floor plan designs, we divided the B2F pipeline in three parts: (3.1) data processing and synthesis, (3.2) training, and (3.3) design.…”
This paper reports a pedagogical experience that incorporates deep learning to design in the context of a recently created course at the Carnegie Mellon University School of Architecture. It analyses an exercise called Bubble2Floor (B2F), where students design floor plans for a multi-story row-house complex. The pipeline for B2F includes a parametric workflow to synthesise an image dataset with pairs of apartment floor plans and corresponding bubble diagrams, a modified Pix2Pix model that maps bubble diagrams to floor plan diagrams, and a computer vision workflow to translate images to the geometric model. In this pedagogical research, we provide a series of observations on challenges faced by students and how they customised different elements of B2F, to address their personal preferences and problem constraints of the housing complex as well as the obstacles from the computational workflow. Based on these observations, we conclude by emphasising the importance of training architects to be active agents in the creation of deep learning workflows and make them accessible for socially relevant and constrained design problems, such as housing.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.