pith. sign in

arxiv: 2607.00015 · v1 · pith:PHQ2GNQ4new · submitted 2026-05-26 · 💻 cs.CY · cs.AI· cs.CV· cs.ET

Towards an automated AI-based framework for floor plan compliance checks for residential buildings

Subash Gautam , Debaditya Acharya , Alexandra Kleeman , Sarah Foster This is my paper

Pith reviewed 2026-07-02 23:26 UTC · model grok-4.3

classification 💻 cs.CY cs.AIcs.CVcs.ET
keywords automated compliance checkingfloor plan analysisAI frameworkbuilding regulationsLLMimage segmentationresidential buildingsdesign standards
0
0 comments X

The pith

An AI framework converts building codes into rules and floor plan images into graphs for automated compliance checks.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper identifies manual review as the main barrier to enforcing apartment design standards on daylight, ventilation, privacy, and space efficiency across thousands of units. It proposes a three-part conceptual framework that uses a large language model to turn regulatory text into executable rules, segments floor plan images into topological graphs of rooms and elements, and then applies the rules to the graphs. If the components work as described, the system would replace case-by-case human checks with a consistent process usable across jurisdictions and building volumes. The authors position this as a way to support large-scale enforcement of existing Australian policies.

Core claim

The paper presents a conceptual framework for automated compliance checking in multi-apartment buildings. A Rule Engine uses an LLM to translate textual building codes into executable, explainable rules. A Data Extraction Engine segments floor plan images into walls, rooms, fixtures, text, and symbols and converts them into a structured building graph with topological relationships. A Compliance Check Engine then evaluates the graph against the LLM-generated rules.

What carries the argument

The three-engine framework: an LLM-based Rule Engine that produces executable rules, a Data Extraction Engine that turns images into topological graphs, and a Compliance Check Engine that applies the rules to the graphs.

If this is right

  • Compliance checking becomes feasible for large numbers of apartments instead of being limited to small samples.
  • The same rule set can be applied consistently across different buildings and jurisdictions.
  • Enforcement of standards for daylight access, ventilation, privacy, and space efficiency can occur at the scale of entire urban developments.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • The graph representation could support additional spatial analyses beyond the listed standards.
  • Pilot deployments on real submitted plans would be needed to quantify error rates on varied drawing styles.
  • The framework might integrate with digital submission portals used by planning authorities.

Load-bearing premise

AI models can accurately translate complex, jurisdiction-specific building codes into correct executable rules and parse diverse floor plan images into accurate topological graphs without substantial human oversight or error.

What would settle it

Apply the framework to a collection of floor plans whose compliance status has already been determined by expert manual review and measure whether the automated results match the manual determinations on a large fraction of cases.

Figures

Figures reproduced from arXiv: 2607.00015 by Alexandra Kleeman, Debaditya Acharya, Sarah Foster, Subash Gautam.

Figure 1
Figure 1. Figure 1: A multi-apartment floorplan from MLStructFP dataset (Pizarro et al., 2023). There is no textual description of the apartments’ [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: Flowchart for Compliance Checking System [PITH_FULL_IMAGE:figures/full_fig_p007_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: A typical architectural floor plan image of a residential apartment (Queensland Housing, 2021). [PITH_FULL_IMAGE:figures/full_fig_p008_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: (a) Advanced text recognition is applied for finding the scale of the image, (b) The segmented geometry as polygon coordinates [PITH_FULL_IMAGE:figures/full_fig_p008_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: Bearing-based aspect extraction from a floor plan. The figure shows how the north arrow, or compass rose, is used to estimate [PITH_FULL_IMAGE:figures/full_fig_p010_5.png] view at source ↗
Figure 6
Figure 6. Figure 6: Diversity in north arrow. However, detecting the north arrow remains challenging because its appearance varies widely across architectural drawings in terms of size, shape, location, and graphical style as shown in [PITH_FULL_IMAGE:figures/full_fig_p010_6.png] view at source ↗
Figure 7
Figure 7. Figure 7: Object detection and segmentation in floor plan image analysis showing two sample apartments from a multi-apartment floor [PITH_FULL_IMAGE:figures/full_fig_p011_7.png] view at source ↗
Figure 8
Figure 8. Figure 8: Data extraction result 12 [PITH_FULL_IMAGE:figures/full_fig_p012_8.png] view at source ↗
Figure 9
Figure 9. Figure 9: (a) Apartment size rule from High Life Study Hooper et al. (2022) (b) & (c) Rule in the compliance document (from (a))) [PITH_FULL_IMAGE:figures/full_fig_p014_9.png] view at source ↗
Figure 10
Figure 10. Figure 10: A sample of results from the compliance engine check where the minimum apartment area of APT_101 Figure 8 is checked [PITH_FULL_IMAGE:figures/full_fig_p015_10.png] view at source ↗
Figure 11
Figure 11. Figure 11: VLM analysis of floor plan image. (a) instruction to vision language model (b) a single-unit apartment (de las Heras et al., [PITH_FULL_IMAGE:figures/full_fig_p018_11.png] view at source ↗
read the original abstract

To improve residents' well-being in Australia's urban areas, governments have introduced policy reforms such as SEPP65, BADS, and SPP7.3 to enhance apartment design quality. These regulations require precise geometric and spatial analysis to evaluate health-related features, including daylight access, natural ventilation, privacy, and space efficiency. However, compliance checking remains challenging due to its manual, time-intensive nature. Additionally, evolving policies limit scalability for large-scale assessments across thousands of apartments. Existing automated floor plan analysis methods are fragmented and typically focus on single apartments, lacking a unified framework for multi-unit compliance checking. This article explores current advancements in automated floor plan analysis, particularly AI-driven approaches, and highlights key challenges in their practical adoption. To address these gaps, a conceptual framework is proposed for automated compliance checking in multi-apartment buildings. A Large Language Model (LLM) is used within a Rule Engine to convert textual building codes into executable, explainable rules. A Data Extraction Engine segments floor plan images into elements such as walls, rooms, fixtures, text, and symbols, and transforms them into a structured building graph with topological relationships. This structured representation is then evaluated by a Compliance Check Engine, which leverages LLM-generated rules for assessment. The proposed framework offers a scalable, consistent, and transparent approach to automated compliance checking across jurisdictions, supporting efficient enforcement of apartment design standards and promoting healthier, higher-density urban development.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

3 major / 2 minor

Summary. The manuscript proposes a conceptual framework for automating compliance checks of multi-apartment residential floor plans against Australian regulations (SEPP65, BADS, SPP7.3). It reviews limitations of existing fragmented methods and describes three components: an LLM-based Rule Engine to translate textual codes into executable rules, a Data Extraction Engine that segments floor-plan images into elements and builds topological graphs, and a Compliance Check Engine that applies the rules to the graphs. The work positions the framework as addressing scalability issues for large-scale assessments.

Significance. If the components could be reliably realized, the framework would address a genuine practical need for consistent enforcement of design standards that affect resident well-being. The integration of LLMs for explainable rule generation and graph-based spatial representation is a coherent architectural idea. However, the manuscript contains no prototype, test cases, accuracy metrics, or feasibility analysis, so any significance remains prospective rather than demonstrated.

major comments (3)
  1. [Abstract] Abstract: the assertion that the framework 'offers a scalable, consistent, and transparent approach to automated compliance checking across jurisdictions' is presented as a delivered outcome, yet the manuscript reports no implementation, no test cases on code-to-rule conversion, and no error analysis for image-to-graph parsing; this claim is therefore unsupported and load-bearing for the paper's central contribution.
  2. [Framework description] Framework description (Rule Engine component): the assumption that an LLM can convert jurisdiction-specific codes into executable rules without misinterpretation or hallucination is stated without any proposed validation protocol, mitigation strategy, or example output, which directly underpins the claimed transparency and cross-jurisdiction consistency.
  3. [Data Extraction Engine] Data Extraction Engine: the transformation of diverse floor-plan images into correct topological graphs is described at a high level with no discussion of handling scale variations, ambiguous symbols, or non-standard drawings, leaving the reliability required for the Compliance Check Engine unaddressed.
minor comments (2)
  1. [Abstract] The abstract and introduction should explicitly label the contribution as a conceptual proposal rather than an operational system to avoid overstating current capabilities.
  2. [Conclusion] A dedicated limitations or future-work subsection would help readers understand the gap between the proposed architecture and practical deployment.

Simulated Author's Rebuttal

3 responses · 1 unresolved

We thank the referee for the constructive comments highlighting the conceptual nature of our framework proposal. We will revise the manuscript to temper claims, clarify scope, and expand discussions on challenges and validation where feasible.

read point-by-point responses

  1. Referee: [Abstract] Abstract: the assertion that the framework 'offers a scalable, consistent, and transparent approach to automated compliance checking across jurisdictions' is presented as a delivered outcome, yet the manuscript reports no implementation, no test cases on code-to-rule conversion, and no error analysis for image-to-graph parsing; this claim is therefore unsupported and load-bearing for the paper's central contribution.

    Authors: We agree the abstract phrasing presents prospective benefits as achieved outcomes. The manuscript is explicitly a conceptual proposal. We will revise the abstract to state that the framework is proposed to offer these qualities, subject to future implementation and empirical validation. revision: yes

  2. Referee: [Framework description] Framework description (Rule Engine component): the assumption that an LLM can convert jurisdiction-specific codes into executable rules without misinterpretation or hallucination is stated without any proposed validation protocol, mitigation strategy, or example output, which directly underpins the claimed transparency and cross-jurisdiction consistency.

    Authors: As a high-level conceptual framework, specific protocols were not detailed. We will add a subsection outlining potential validation approaches (e.g., expert review of LLM outputs, prompt engineering for consistency, and cross-jurisdiction testing) to support transparency claims. Concrete examples and full mitigation strategies require implementation, which is outside the paper's scope. revision: partial

  3. Referee: [Data Extraction Engine] Data Extraction Engine: the transformation of diverse floor-plan images into correct topological graphs is described at a high level with no discussion of handling scale variations, ambiguous symbols, or non-standard drawings, leaving the reliability required for the Compliance Check Engine unaddressed.

    Authors: We will expand the Data Extraction Engine section to explicitly discuss these challenges and outline mitigation strategies, such as multi-resolution processing for scale variations, symbol libraries with uncertainty handling, and hybrid CV-LLM approaches for non-standard drawings, to better ground the reliability assumptions. revision: yes

standing simulated objections not resolved
  • Provision of a prototype, test cases, accuracy metrics, or feasibility analysis, as the manuscript proposes a conceptual framework without implementation or empirical evaluation.

Circularity Check

0 steps flagged

No circularity: descriptive conceptual proposal without derivations or self-referential logic

full rationale

The paper proposes a high-level architecture for an AI-based compliance checking system using LLMs for rule extraction and image segmentation for graph generation. It contains no equations, no fitted parameters, no 'predictions' of any kind, and no self-citations that support load-bearing claims. The central statements are forward-looking descriptions of intended components rather than derivations that reduce to their own inputs. This matches the default expectation of a non-circular descriptive manuscript.

Axiom & Free-Parameter Ledger

0 free parameters · 2 axioms · 3 invented entities

The central claim rests on untested assumptions about LLM and CV performance in this domain plus newly introduced framework components without independent validation or evidence.

axioms (2)
  • domain assumption Large Language Models can accurately translate building code text into executable, explainable rules.
    Invoked for the Rule Engine without supporting evidence or testing.
  • domain assumption Floor plan images can be segmented into accurate structured building graphs with topological relationships.
    Invoked for the Data Extraction Engine without supporting evidence or testing.
invented entities (3)
  • Rule Engine (LLM-based) no independent evidence
    purpose: Convert textual building codes into executable rules
    New component proposed in the framework; no independent evidence of reliability for this use case.
  • Data Extraction Engine no independent evidence
    purpose: Segment floor plan images into elements and transform into building graph
    New component proposed in the framework; no independent evidence of reliability for this use case.
  • Compliance Check Engine no independent evidence
    purpose: Evaluate building graph against LLM-generated rules
    New component proposed in the framework; no independent evidence of reliability for this use case.

pith-pipeline@v0.9.1-grok · 5798 in / 1410 out tokens · 29014 ms · 2026-07-02T23:26:54.838480+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

128 extracted references · 61 canonical work pages · 5 internal anchors

  1. [1]

    and Martel, A

    Allouf, D. and Martel, A. and March, A. , title =. Environment and Planning B: Urban Analytics and City Science , volume =. doi:10.1177/2399808318825273 , year =

    work page doi:10.1177/2399808318825273

  2. [2]

    and Touchie, M

    Andargie, M.S. and Touchie, M. and O’Brian, W. , title =. Building Acoustics , volume =. doi:10.1177/1351010X21993742 , year =

    work page doi:10.1177/1351010x21993742

  3. [3]

    2022 , type =

    2021 Census count includes Australians living on wheels and water, but most of us still firmly on land , publisher =. 2022 , type =

    2021

  4. [4]

    and Wolke, G

    Babisch, W. and Wolke, G. and Heinrich, J. and Straff, W. , title =. ENVIRONMENTAL RESEARCH , volume =. doi:10.1016/j.envres.2014.05.007 , year =

    work page doi:10.1016/j.envres.2014.05.007 2014

  5. [5]

    Brown, M. J. and Jacobs, D. E. , title =. Public Health Reports , volume =. 2011 , type =

    2011

  6. [6]

    and Judd, S

    Easthope, H. and Judd, S. , title =. 2010 , type =

    2010

  7. [7]

    Evans, G. W. and Lepore, S. J. and Schroeder, A. , title =. JOURNAL OF PERSONALITY AND SOCIAL PSYCHOLOGY , volume =. 1996 , type =

    1996

  8. [8]

    and Wells, N.M

    Evans, G.W. and Wells, N.M. and Moch, A. , title =. Journal of Social Issues , volume =. 2003 , type =

    2003

  9. [9]

    and Hooper, P

    Foster, S. and Hooper, P. and Duckworth, A. and Bolleter, J. , title =. Building and Environment , volume =. doi:10.1016/j.buildenv.2021.108493 , year =

    work page doi:10.1016/j.buildenv.2021.108493 2021

  10. [10]

    and Hooper, P

    Foster, S. and Hooper, P. and Turrell, G. and Maitland, C. and Giles-Corti, B. and Kleeman, A. , title =. Social Science & Medicine - Population Health , volume =. doi:10.1016/j.ssmph.2022.101301 , year =

    work page doi:10.1016/j.ssmph.2022.101301 2022

  11. [11]

    and Maitland, C

    Foster, S. and Maitland, C. and Hooper, P. and Bolleter, J. and Duckworth-Smith, A. and Giles-Corti, B. and Arundel, J. , title =. BMJ Open , volume =. doi:10.1136/bmjopen-2019-029220 , year =

    work page doi:10.1136/bmjopen-2019-029220 2019

  12. [12]

    and Kleeman, A

    Giles-Corti, B. and Kleeman, A. and Foster, S. , title =. 2015 , type =

    2015

  13. [13]

    and Bolleter, J

    Hooper, P. and Bolleter, J. and Edwards, N. and Kleeman, A. and Duckworth, A. and Foster, S. , title =. MethodsX , volume =. doi:10.1016/j.mex.2022.101810 , year =

    work page doi:10.1016/j.mex.2022.101810 2022

  14. [14]

    and Kleeman, A

    Hooper, P. and Kleeman, A. and Edwards, N. and Bolleter, J. and Foster, S. , title =. The Lancet Regional Health – Western Pacific , volume =. doi:10.1016/j.lanwpc.2023.100807 , year =

    work page doi:10.1016/j.lanwpc.2023.100807 2023

  15. [15]

    , title =

    Kaplan, R. , title =. Environment and Behavior , volume =. 2001 , type =

    2001

  16. [16]

    and Foster, S

    Kleeman, A. and Foster, S. , title =. Journal of Environmental Psychology , volume =. doi:10.1016/j.jenvp.2023.102056 , year =

    work page doi:10.1016/j.jenvp.2023.102056 2023

  17. [17]

    and Foster, S

    Kleeman, A. and Foster, S. , title =. Cities & Health , volume =. doi:10.1080/23748834.2023.2294639 , year =

    work page doi:10.1080/23748834.2023.2294639 2023

  18. [18]

    and Giles-Corti, B

    Kleeman, A. and Giles-Corti, B. and Gunn, L. and Hooper, P. and Foster, S. , title =. Cities , volume =. doi:10.1016/j.cities.2022.104126 , year =

    work page doi:10.1016/j.cities.2022.104126 2022

  19. [19]

    and Hooper, P

    Kleeman, A. and Hooper, P. and Giles-Corti, B. and Foster, S. , title =. Cities , volume =. doi:10.1016/j.cities.2022.104043 , year =

    work page doi:10.1016/j.cities.2022.104043 2022

  20. [20]

    Lloyd, E. L. and McCormack, C. and McKeever, M. and Syme, M. , title =. JOURNAL OF EPIDEMIOLOGY AND COMMUNITY HEALTH , volume =. doi:10.1136/jech.2007.067835 , year =

    work page doi:10.1136/jech.2007.067835 2007

  21. [21]

    , title =

    Mould, P. , title =. Architecture Australia , volume =. 2011 , type =

    2011

  22. [22]

    2024 , type =

    State of the Housing System 2024 , publisher =. 2024 , type =

    2024

  23. [23]

    2015 , type =

    Apartment design guide – Tools for improving the design of residential apartment development , publisher =. 2015 , type =

    2015

  24. [24]

    2021 , type =

    Apartment Design Guidelines for Victoria , publisher =. 2021 , type =

    2021

  25. [25]

    2024 , type =

    Residential Design Codes – Volume 2 , publisher =. 2024 , type =

    2024

  26. [26]

    Wong, N. H. and Huang, B. , title =. BUILDING AND ENVIRONMENT , volume =. doi:10.1016/j.buildenv.2004.01.024 , year =

    work page doi:10.1016/j.buildenv.2004.01.024 2004

  27. [27]

    Advances in neural information processing systems , volume=

    Chain-of-thought prompting elicits reasoning in large language models , author=. Advances in neural information processing systems , volume=

  28. [28]

    ACM computing surveys (CSUR) , volume=

    A deep journey into super-resolution: A survey , author=. ACM computing surveys (CSUR) , volume=. 2020 , publisher=

    2020

  29. [29]

    Proceedings of the 12th ACM International Conference on Systems for Energy-Efficient Buildings, Cities, and Transportation , pages=

    A Vision-Language Model Agent for building code compliance , author=. Proceedings of the 12th ACM International Conference on Systems for Energy-Efficient Buildings, Cities, and Transportation , pages=

  30. [30]

    arXiv preprint arXiv:2510.02634 , year=

    Automatic Building Code Review: A Case Study , author=. arXiv preprint arXiv:2510.02634 , year=

    work page arXiv

  31. [31]

    Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing , pages=

    Topviewrs: Vision-language models as top-view spatial reasoners , author=. Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing , pages=

    2024

  32. [32]

    Buildings , volume=

    Automated building information modeling compliance check through a large language model combined with deep learning and ontology , author=. Buildings , volume=. 2024 , publisher=

    2024

  33. [33]

    Automation in Construction , volume=

    Vision-language model-based intelligent assistant for onsite construction safety inspection , author=. Automation in Construction , volume=. 2026 , publisher=

    2026

  34. [34]

    arXiv preprint arXiv:2409.12842 , year=

    Vision language models can parse floor plan maps , author=. arXiv preprint arXiv:2409.12842 , year=

    work page arXiv

  35. [35]

    arXiv preprint arXiv:2602.06507 , year=

    FloorplanVLM: A Vision-Language Model for Floorplan Vectorization , author=. arXiv preprint arXiv:2602.06507 , year=

    work page arXiv

  36. [36]

    Advanced Engineering Informatics , volume=

    Natural language generation and deep learning for intelligent building codes , author=. Advanced Engineering Informatics , volume=. 2022 , publisher=

    2022

  37. [37]

    2016 , issn =

    Graph theory based representation of building information models for access control applications , journal =. 2016 , issn =. doi:https://doi.org/10.1016/j.autcon.2016.04.001 , url =

    work page doi:10.1016/j.autcon.2016.04.001 2016

  38. [38]

    Journal of Management in Engineering , volume=

    Enriched construction regulation inquiry responses: A hybrid search approach for large language models , author=. Journal of Management in Engineering , volume=. 2025 , publisher=

    2025

  39. [39]

    Automation in Construction , volume=

    Prompt-based automation of building code information transformation for compliance checking , author=. Automation in Construction , volume=. 2024 , publisher=

    2024

  40. [40]

    Scientific data , volume=

    CODE-ACCORD: A Corpus of building regulatory data for rule generation towards automatic compliance checking , author=. Scientific data , volume=. 2025 , publisher=

    2025

  41. [41]

    IEEE Access , volume=

    Human-in-the-loop learning with LLMs for efficient RASE tagging in building compliance regulations , author=. IEEE Access , volume=. 2024 , publisher=

    2024

  42. [42]

    IOP Conference Series: Earth and Environmental Science , volume=

    Neural semantic parsing of building regulations for compliance checking , author=. IOP Conference Series: Earth and Environmental Science , volume=. 2022 , organization=

    2022

  43. [43]

    Advanced Engineering Informatics , volume=

    Intermediate representations to improve the semantic parsing of building regulations , author=. Advanced Engineering Informatics , volume=. 2024 , publisher=

    2024

  44. [44]

    arXiv preprint arXiv:2509.17283 , year=

    Automated Facility Enumeration for Building Compliance Checking using Door Detection and Large Language Models , author=. arXiv preprint arXiv:2509.17283 , year=

    work page arXiv

  45. [45]

    RC-Net: Row&Column Net with Text Feature for Deep Parsing Floor Plan Images , author=

  46. [46]

    Advanced Engineering Informatics , volume=

    Automated fire risk assessment and mitigation in building blueprints using computer vision and deep generative models , author=. Advanced Engineering Informatics , volume=. 2024 , publisher=

    2024

  47. [47]

    European Conference on Computer Vision , pages=

    MSD: A Benchmark Dataset for Floor Plan Generation of Building Complexes , author=. European Conference on Computer Vision , pages=. 2024 , organization=

    2024

  48. [48]

    Proceedings of the Asian Conference on Computer Vision , pages=

    Vectorizing building blueprints , author=. Proceedings of the Asian Conference on Computer Vision , pages=

  49. [49]

    2018 11th International Congress on Image and Signal Processing, BioMedical Engineering and Informatics (CISP-BMEI) , pages=

    Semantic segmentation in architectural floor plans for detecting walls and doors , author=. 2018 11th International Congress on Image and Signal Processing, BioMedical Engineering and Informatics (CISP-BMEI) , pages=. 2018 , organization=

    2018

  50. [50]

    Journal of Computing in Civil Engineering , volume=

    Hierarchical Intersection over Union Method: A Comprehensive Evaluation for Floor Plan Vectorization , author=. Journal of Computing in Civil Engineering , volume=. 2025 , publisher=

    2025

  51. [51]

    International Journal on Document Analysis and Recognition (IJDAR) , volume=

    Automatic floor plan analysis using a boundary attention-based deep network , author=. International Journal on Document Analysis and Recognition (IJDAR) , volume=. 2025 , publisher=

    2025

  52. [52]

    ISPRS International Journal of Geo-Information , volume=

    Automatic extraction of indoor spatial information from floor plan image: A patch-based deep learning methodology application on large-scale complex buildings , author=. ISPRS International Journal of Geo-Information , volume=. 2021 , publisher=

    2021

  53. [53]

    IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing , volume=

    Automated semantics and topology representation of residential-building space using floor-plan raster maps , author=. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing , volume=. 2022 , publisher=

    2022

  54. [54]

    2024 8th SLAAI International Conference on Artificial Intelligence (SLAAI-ICAI) , pages=

    A Promptable Segmentation Approach to Automatic Floor Plan Analysis using Vision Transformers , author=. 2024 8th SLAAI International Conference on Artificial Intelligence (SLAAI-ICAI) , pages=. 2024 , organization=

    2024

  55. [55]

    Semantic and instance segmentation of room features in floor plans using Mask R-CNN , author=

  56. [56]

    European Conference on Computer Vision , pages=

    PolyRoom: Room-aware transformer for floorplan reconstruction , author=. European Conference on Computer Vision , pages=. 2024 , organization=

    2024

  57. [57]

    Proceedings of the 2021 4th International Conference on Artificial Intelligence and Pattern Recognition , pages=

    Floor plan semantic segmentation using deep learning with boundary attention aggregated mechanism , author=. Proceedings of the 2021 4th International Conference on Artificial Intelligence and Pattern Recognition , pages=

    2021

  58. [58]

    Progress in Engineering Science , pages=

    Generative artificial intelligence (AI) in built environment design and planning--A state-of-the-art review , author=. Progress in Engineering Science , pages=. 2024 , publisher=

    2024

  59. [59]

    Journal of Infrastructure Intelligence and Resilience , volume=

    Few-shot learning with large foundation models for automated segmentation and accessibility analysis in architectural floor plans , author=. Journal of Infrastructure Intelligence and Resilience , volume=. 2025 , publisher=

    2025

  60. [60]

    2017 international conference on research and innovation in information systems (icriis) , pages=

    A review on BIM-based automated code compliance checking system , author=. 2017 international conference on research and innovation in information systems (icriis) , pages=. 2017 , organization=

    2017

  61. [61]

    Journal of Building Engineering , volume=

    Generating BIM model from structural and architectural plans using Artificial Intelligence , author=. Journal of Building Engineering , volume=. 2023 , publisher=

    2023

  62. [62]

    Automation in Construction , volume=

    Benchmarking methods for classifying space functions and access elements in multi-unit apartment buildings , author=. Automation in Construction , volume=. 2025 , publisher=

    2025

  63. [63]

    Automation in Construction , volume=

    Comprehensive floor plan vectorization with sparse point set representation , author=. Automation in Construction , volume=. 2025 , publisher=

    2025

  64. [64]

    Automation in Construction , volume=

    Semantic floorplan segmentation using self-constructing graph networks , author=. Automation in Construction , volume=. 2024 , publisher=

    2024

  65. [65]

    Automation in Construction , volume=

    Automatic floor plan analysis and recognition , author=. Automation in Construction , volume=. 2022 , publisher=

    2022

  66. [66]

    Madireddy, Soumya and Gao, Lu and Din, Zia and Kim, Kinam and Senouci, Ahmed and Han, Zhe and Zhang, Yunpeng , month = jun, year =. Large. doi:10.48550/arXiv.2506.20551 , urldate =

    work page doi:10.48550/arxiv.2506.20551

  67. [67]

    Improved

    Ahmed, Sheraz and Liwicki, Marcus and Weber, Markus and Dengel, Andreas , month = sep, year =. Improved. 2011. doi:10.1109/ICDAR.2011.177 , urldate =

    work page doi:10.1109/icdar.2011.177 2011

  68. [68]

    Journal of Electronic Imaging , author =

    Semantic segmentation of multiclass walls in complex architectural floor plan image , volume =. Journal of Electronic Imaging , author =. 2024 , note =. doi:10.1117/1.JEI.33.4.043040 , number =

    work page doi:10.1117/1.jei.33.4.043040 2024

  69. [69]

    Parsing floor plan images , url =

    Dodge, Samuel and Xu, Jiu and Stenger, Björn , month = may, year =. Parsing floor plan images , url =. 2017. doi:10.23919/MVA.2017.7986875 , urldate =

    work page doi:10.23919/mva.2017.7986875 2017

  70. [70]

    Automation in Construction , author =

    Large-scale multi-unit floor plan dataset for architectural plan analysis and recognition , volume =. Automation in Construction , author =. 2023 , keywords =. doi:10.1016/j.autcon.2023.105132 , urldate =

    work page doi:10.1016/j.autcon.2023.105132 2023

  71. [71]

    International Journal on Document Analysis and Recognition (IJDAR) , author =

    Generation of synthetic documents for performance evaluation of symbol recognition & spotting systems , volume =. International Journal on Document Analysis and Recognition (IJDAR) , author =. 2010 , keywords =. doi:10.1007/s10032-010-0120-x , language =

    work page doi:10.1007/s10032-010-0120-x 2010

  72. [73]

    Xu, Zhongguo and Jha, Naresh and Mehadi, Syed and Mandal, Mrinal , month = jun, year =. Deep. Proceedings of the 2023. doi:10.1145/3596286.3596295 , abstract =

    work page doi:10.1145/3596286.3596295 2023

  73. [74]

    Information Sciences , author =

    Data-driven floor plan understanding in rural residential buildings via deep recognition , volume =. Information Sciences , author =. 2021 , keywords =. doi:10.1016/j.ins.2021.03.032 , abstract =

    work page doi:10.1016/j.ins.2021.03.032 2021

  74. [75]

    Automation in Construction , author =

    Automatic floor plan analysis:. Automation in Construction , author =. 2025 , keywords =. doi:10.1016/j.autcon.2025.106378 , abstract =

    work page doi:10.1016/j.autcon.2025.106378 2025

  75. [76]

    Advanced Engineering Informatics , author =

    Automated fire risk assessment and mitigation in building blueprints using computer vision and deep generative models , volume =. Advanced Engineering Informatics , author =. 2024 , keywords =. doi:10.1016/j.aei.2024.102614 , abstract =

    work page doi:10.1016/j.aei.2024.102614 2024

  76. [77]

    Rapid Prototyping Journal , author =

    A state-of-the-art digital factory integrating digital twin for laser additive and subtractive manufacturing processes , volume =. Rapid Prototyping Journal , author =. 2023 , pages =. doi:10.1108/RPJ-03-2023-0113 , abstract =

    work page doi:10.1108/rpj-03-2023-0113 2023

  77. [78]

    International Journal on Document Analysis and Recognition (IJDAR) , author =

    A. International Journal on Document Analysis and Recognition (IJDAR) , author =. 2025 , keywords =. doi:10.1007/s10032-025-00528-8 , abstract =

    work page doi:10.1007/s10032-025-00528-8 2025

  78. [79]

    2017 14th

    Sharma, Divya and Gupta, Nitin and Chattopadhyay, Chiranjoy and Mehta, Sameep , month = nov, year =. 2017 14th. doi:10.1109/ICDAR.2017.76 , abstract =

    work page doi:10.1109/icdar.2017.76 2017

  79. [81]

    Automation in Construction , author =

    Multiscale object detection on complex architectural floor plans , volume =. Automation in Construction , author =. 2024 , keywords =. doi:10.1016/j.autcon.2024.105486 , abstract =

    work page doi:10.1016/j.autcon.2024.105486 2024

  80. [82]

    Automation in Construction , author =

    Comprehensive floor plan vectorization with sparse point set representation , volume =. Automation in Construction , author =. 2025 , keywords =. doi:10.1016/j.autcon.2025.106023 , abstract =

    work page doi:10.1016/j.autcon.2025.106023 2025

Showing first 80 references.