Research map / Image generation / GANs
Roof-GAN: Learning To Generate Roof Geometry and Relations for Residential Houses
Yiming Qian, Hao Zhang, Yasutaka Furukawa
Check your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| Add generated pairwise relationships, a relationship discriminator, and differentiable relationship-aware geometry processing (ablation) | Roof-GAN · Residential roof generation test set | RMMD: 6.61 -> 6.20 | Improved |
| Add differentiable vectorization and relationship enforcement during training (ablation) | Roof-GAN · Residential roof generation test set | RMMD: 6.48 -> 6.20 | Improved |
| Encode pairwise relationships in the serialized representation | PQ-Net · Residential roof generation test set | RMMD: 7.80 -> 6.97 | Improved |
| Encode pairwise relationships in the serialized representation | PQ-Net · Residential roof generation test set | FID: 12.0 -> 13.5 | Hurt |
| Use Roof-GAN's raster geometry representation instead of House-GAN's vectorized facet-angle and primitive-type representation | House-GAN · Residential roof generation test set | FID: 18.4 -> 10.6 | Improved |
| Use Roof-GAN's raster geometry representation instead of House-GAN's vectorized facet-angle and primitive-type representation | House-GAN · Residential roof generation test set | RMMD: 7.43 -> 6.61 | Improved |
| Generate with Roof-GAN rather than PQ-Net | Roof-GAN · Cross-validation, 3 primitives | RMMD: 10.4 -> 7.5 | Improved |
| Generate with Roof-GAN rather than House-GAN | Roof-GAN · Cross-validation, 3 primitives | RMMD: 8.5 -> 7.5 | Improved |
| Generate with Roof-GAN rather than PQ-Net | Roof-GAN · Cross-validation, 4 primitives | FID: 14.6 -> 13.8 | Improved |
| Generate with Roof-GAN rather than PQ-Net | Roof-GAN · Cross-validation, 4 primitives | RMMD: 12.9 -> 10.9 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- House-GAN — House-GAN: Relational Generative Adversarial Networks for Graph-constrained House Layout Generation (ECCV 2020) · better on 6 of 6 shared results
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- House-GAN: Relational Generative Adversarial Networks for Graph-constrained House Layout Generation (ECCV 2020 · similarity 0.90)
- House-GAN++: Generative Adversarial Layout Refinement Network towards Intelligent Computational Agent for Professional Architects (CVPR 2021 · similarity 0.88)
- Graph Transformer GANs for Graph-Constrained House Generation (CVPR 2023 · similarity 0.88)
- Intelligent Home 3D: Automatic 3D-House Design From Linguistic Descriptions Only (CVPR 2020 · similarity 0.87)
- Graph Transformer GANs With Graph Masked Modeling for Architectural Layout Generation (TPAMI 2024 · similarity 0.86)
- GSDiff: Synthesizing Vector Floorplans via Geometry-enhanced Structural Graph Generation (AAAI 2025 · similarity 0.84)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
