Research map / 3D vision / Depth and stereo
3D Building Reconstruction From Monocular Remote Sensing Images
Weijia Li, Lingxuan Meng, Jinwang Wang, Conghui He, Gui-Song Xia, Dahua Lin
Read the paperCheck 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 offset field A prediction (warp roofs to footprints by average offset) (ablation) | MTBR-Net baseline · footprint, in-domain | F1-score: 66.04 -> 66.79 | Improved |
| add offset field A prediction (ablation) | MTBR-Net baseline · footprint, out-domain | F1-score: 47.65 -> 48.49 | Improved |
| add offset-based feature warping module (offset field B) (ablation) | MTBR-Net with offset field · footprint, in-domain | F1-score: 66.79 -> 67.44 | Improved |
| add offset-based feature warping module (offset field B) (ablation) | MTBR-Net with offset field · footprint, out-domain | F1-score: 48.49 -> 50.04 | Improved |
| add prior knowledge based 3D model optimization (ablation) | MTBR-Net with feature warping · footprint, in-domain | F1-score: 67.44 -> 68.58 | Improved |
| add prior knowledge based 3D model optimization (ablation) | MTBR-Net with feature warping · footprint, out-domain | F1-score: 50.04 -> 51.98 | Improved |
| add feature warping to offset field (ablation) | MTBR-Net · height EPE, in-domain | EPE (pixels): 5.26 -> 5.17 | Improved |
| add 3D model optimization (ablation) | MTBR-Net · height EPE, in-domain | EPE (pixels): 5.17 -> 4.88 | Improved |
| add 3D model optimization (ablation) | MTBR-Net · height EPE, out-domain | EPE (pixels): 10.21 -> 9.59 | Improved |
| add offset field prediction (height estimation) (ablation) | MTBR-Net baseline · height EPE, out-domain | EPE (pixels): 10.45 -> 10.21 | Improved |
| train interrelated multi-task roof segmentation instead of single-task HR-Net | MTBR-Net · roof segmentation, in/out-domain | F1-score: improves F1 by 4.6% and 10.6% | Improved |
| replace Christie et al. with multi-task MTBR-Net for height | MTBR-Net · DFC19 actual height RMSE | RMSE (m): 16.87 -> 9.57 | Improved |
Closest papers by meaning
- 3D Building Reconstruction from Monocular Remote Sensing Images with Multi-level Supervisions (CVPR 2024 · similarity 0.90)
- SynRS3D: A Synthetic Dataset for Global 3D Semantic Understanding from Monocular Remote Sensing Imagery (NeurIPS 2024 · similarity 0.85)
- BWFormer: Building Wireframe Reconstruction from Airborne LiDAR Point Cloud with Transformer (CVPR 2025 · similarity 0.83)
- PBWR: Parametric-Building-Wireframe Reconstruction from Aerial LiDAR Point Clouds (CVPR 2024 · similarity 0.83)
- Aerial Lifting: Neural Urban Semantic and Building Instance Lifting from Aerial Imagery (CVPR 2024 · similarity 0.82)
- MP-SfM: Monocular Surface Priors for Robust Structure-from-Motion (CVPR 2025 · similarity 0.82)
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.
