Saved in:
| Main Authors: | , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2026
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2602.07100 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866918326084567040 |
|---|---|
| author | Xiong, Biao Peng, Zhen Wang, Ping Liu, Qiegen Zhong, Xian |
| author_facet | Xiong, Biao Peng, Zhen Wang, Ping Liu, Qiegen Zhong, Xian |
| contents | Automated floorplan generation aims to improve design quality, architectural efficiency, and sustainability by jointly modeling global spatial organization and precise geometric detail. However, existing approaches operate in raster space and rely on post hoc vectorization, which introduces structural inconsistencies and hinders end-to-end learning. Motivated by compositional spatial reasoning, we propose TLC-Plan, a hierarchical generative model that directly synthesizes vector floorplans from input boundaries, aligning with human architectural workflows based on modular and reusable patterns. TLC-Plan employs a two-level VQ-VAE to encode global layouts as semantically labeled room bounding boxes and to refine local geometries using polygon-level codes. This hierarchy is unified in a CodeTree representation, while an autoregressive transformer samples codes conditioned on the boundary to generate diverse and topologically valid designs, without requiring explicit room topology or dimensional priors. Extensive experiments show state-of-the-art performance on RPLAN dataset (FID = 1.84, MSE = 2.06) and leading results on LIFULL dataset. The proposed framework advances constraint-aware and scalable vector floorplan generation for real-world architectural applications. Source code and trained models are released at https://github.com/rosolose/TLC-PLAN. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2602_07100 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | TLC-Plan: A Two-Level Codebook Based Network for End-to-End Vector Floorplan Generation Xiong, Biao Peng, Zhen Wang, Ping Liu, Qiegen Zhong, Xian Computer Vision and Pattern Recognition Automated floorplan generation aims to improve design quality, architectural efficiency, and sustainability by jointly modeling global spatial organization and precise geometric detail. However, existing approaches operate in raster space and rely on post hoc vectorization, which introduces structural inconsistencies and hinders end-to-end learning. Motivated by compositional spatial reasoning, we propose TLC-Plan, a hierarchical generative model that directly synthesizes vector floorplans from input boundaries, aligning with human architectural workflows based on modular and reusable patterns. TLC-Plan employs a two-level VQ-VAE to encode global layouts as semantically labeled room bounding boxes and to refine local geometries using polygon-level codes. This hierarchy is unified in a CodeTree representation, while an autoregressive transformer samples codes conditioned on the boundary to generate diverse and topologically valid designs, without requiring explicit room topology or dimensional priors. Extensive experiments show state-of-the-art performance on RPLAN dataset (FID = 1.84, MSE = 2.06) and leading results on LIFULL dataset. The proposed framework advances constraint-aware and scalable vector floorplan generation for real-world architectural applications. Source code and trained models are released at https://github.com/rosolose/TLC-PLAN. |
| title | TLC-Plan: A Two-Level Codebook Based Network for End-to-End Vector Floorplan Generation |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2602.07100 |