Skip to content

vendkura/akiya-3d

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

3 Commits
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Akiya-3D: Automatic Conversion of 2D Japanese Floorplans to 3D Models

🚧 Preliminary Documentation — This README provides a research overview. Comprehensive documentation including reproducibility instructions, dataset specifications, and detailed methodology will be released upon project completion.

Abstract

This repository presents an end-to-end pipeline for automatically converting 2D Japanese floorplan images into 3D architectural models. Unlike existing approaches (CubiCasa5k, MMDetection-based methods) that target Western architectural conventions, this work addresses the unique characteristics of traditional Japanese residential architecture—including tatami rooms, DK/LDK layouts, fusuma sliding panels, and engawa verandas.

The pipeline combines semantic segmentation using U-Net with ResNet34 encoder (achieving 53.5% mIoU on a custom 13-class taxonomy) with geometric post-processing and wall-based 3D extrusion to produce "dollhouse style" navigable models.

Developed in partnership with akiya2.com to address the visualization needs of foreign buyers acquiring akiya (abandoned traditional Japanese houses).

Repository Structure

Branch Description
main Project documentation and overview
u-net_implementation Core segmentation pipeline — U-Net with ResNet34 encoder, training scripts, inference
3d-pipeline-integration 3D reconstruction — Wall extrusion, boundary extraction, OBJ generation
cubicasa-original-implementation Baseline experiments with CubiCasa5k approach
mmdetection-implementation Baseline experiments with MMDetection framework
media_documentation Figures, visualizations, and supplementary materials

Methodology

Semantic Segmentation

The segmentation model employs a U-Net architecture with ResNet34 encoder pretrained on ImageNet. Training configuration:

  • Input resolution: 512×512
  • Optimizer: Adam (lr=1e-4)
  • Batch size: 4
  • Augmentation: rotation, flipping, elastic deformation, color jittering

Taxonomy Design

A custom 13-class taxonomy was developed for Japanese residential architecture:

Category Classes
Living spaces LDK, DK, Bedroom, Japanese-style room (和室)
Functional Bathroom, Toilet, Kitchen, Entrance (玄関), Storage
Structural Wall, Door, Window
Exterior Balcony/Veranda

Key finding: Taxonomy consolidation (16 → 13 classes) yielded greater mIoU improvement than doubling dataset size, demonstrating the importance of label design in low-resource specialized domains.

3D Reconstruction

The reconstruction pipeline uses wall-based extrusion rather than room-based extrusion:

Input Mask → Boundary Extraction → Morphological Filtering → Wall Extrusion (12cm) → OBJ Export

This approach produces architecturally realistic models where walls are solid geometry and rooms are walkable voids.

Experimental Results

Segmentation Performance Progression

Experiment mIoU Δ Configuration
E1: Baseline 17.7% 16 classes, 50 images, light augmentation
E2: Augmentation 19.3% +1.6 Heavy augmentation strategy
E3: Taxonomy 26.0% +6.7 Consolidated to 13 classes
E4: Data scaling 42.2% +16.2 Extended to 100 images
E5: Fine-tuning 53.5% +11.3 Extended training, hyperparameter optimization

Baseline Comparisons

Experiments with Western-focused approaches revealed significant domain gap:

  • CubiCasa5k: Pretrained models failed on Japanese architectural elements
  • MMDetection: Required extensive adaptation for non-Western room types

These negative results validated the need for a domain-specific approach.

Pipeline Architecture

┌─────────────────┐     ┌──────────────────┐     ┌─────────────────┐
│  Input Image    │────▶│  U-Net Encoder   │────▶│  Segmentation   │
│  (Japanese FP)  │     │  (ResNet34)      │     │  Mask (13 cls)  │
└─────────────────┘     └──────────────────┘     └────────┬────────┘
                                                          │
                                                          ▼
┌─────────────────┐     ┌──────────────────┐     ┌─────────────────┐
│  3D OBJ Model   │◀────│  Wall Extrusion  │◀────│  Boundary       │
│  (Dollhouse)    │     │  (12cm walls)    │     │  Extraction     │
└─────────────────┘     └──────────────────┘     └─────────────────┘

Requirements

  • Python 3.9+
  • PyTorch 2.0+
  • CUDA-capable GPU (12GB+ VRAM recommended)
  • Dependencies: segmentation-models-pytorch, albumentations, OpenCV, trimesh

Detailed requirements and installation instructions will be provided in branch-specific documentation.

Dataset

The model was trained on a custom dataset of ~100 manually annotated Japanese floorplan images sourced through partnership with akiya2.com. Annotations were created using Label Studio following the 13-class taxonomy.

Dataset availability details to be announced.

Acknowledgments

  • akiya2.com — Industry partnership and data provision

License

To be specified upon release.


Full documentation, pretrained weights, and reproducibility materials forthcoming.

About

End-to-end pipeline for automatically converting Japanese floorplan images into 3D architectural models using U-Net semantic segmentation and geometric post-processing. Built to address visualization needs for foreign buyers of akiya (abandoned Japanese houses).

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors