Floor Plan Reconstruction from Sparse Views: Combining Graph Neural Network with Constrained Diffusion
Abstract
We address the challenging problem of floor plan reconstruction from sparse views and a room-connectivity graph. As a first stage, we construct a flexible graph-structure unifying the connectivity graph and the sparse observed data. Using our Graph Neural Network architecture, we can then refine the available information and predict unobserved room properties. In a second step, we introduce a Constrained Diffusion Model to reconstruct consistent floor plan matching the available information, despite of its sparsity. More precisely, we use a Cross-Attention mechanism armed with shape descriptors to guarantee that the generated floor plan reflects both the input room connectivity and the geometry observed in the sparse views.
Cite
Text
Gueze et al. "Floor Plan Reconstruction from Sparse Views: Combining Graph Neural Network with Constrained Diffusion." IEEE/CVF International Conference on Computer Vision Workshops, 2023. doi:10.1109/ICCVW60793.2023.00173Markdown
[Gueze et al. "Floor Plan Reconstruction from Sparse Views: Combining Graph Neural Network with Constrained Diffusion." IEEE/CVF International Conference on Computer Vision Workshops, 2023.](https://mlanthology.org/iccvw/2023/gueze2023iccvw-floor/) doi:10.1109/ICCVW60793.2023.00173BibTeX
@inproceedings{gueze2023iccvw-floor,
title = {{Floor Plan Reconstruction from Sparse Views: Combining Graph Neural Network with Constrained Diffusion}},
author = {Gueze, Arnaud and Ospici, Matthieu and Rohmer, Damien and Cani, Marie-Paule},
booktitle = {IEEE/CVF International Conference on Computer Vision Workshops},
year = {2023},
pages = {1575-1584},
doi = {10.1109/ICCVW60793.2023.00173},
url = {https://mlanthology.org/iccvw/2023/gueze2023iccvw-floor/}
}