Research map / 3D vision / Face and shape
Inner-Outer Aware Reconstruction Model for Monocular 3D Scene Reconstruction
Yu-Kun Qiu, Guohao Xu, Wei-Shi Zheng
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 separate occupancy branches (ablation) | IOAR basic model · ScanNet (VoRTX protocol) | F-score: 0.641 -> 0.654 | Improved |
| add inner-outer aware coarse-to-fine strategy and loss (ablation) | IOAR basic model · ScanNet (VoRTX protocol) | Prec: 0.708 -> 0.743 | Improved |
| add inner-outer aware strategy and loss (ablation) | IOAR basic model · ScanNet (VoRTX protocol) | F-score: 0.641 -> 0.652 | Improved |
| add inner-outer aware strategy and loss (ablation) | IOAR basic model · ScanNet (VoRTX protocol) | Recall: 0.588 -> 0.583 | Hurt |
| combine separate branches with IOA (full IOAR) (ablation) | IOAR · ScanNet (VoRTX protocol) | F-score: 0.654 -> 0.663 | Improved |
| replace VoRTX with IOAR | 3D scene reconstruction · ScanNet (VoRTX protocol) | F-score: 0.641 -> 0.663 | Improved |
| replace VoRTX with IOAR | 3D scene reconstruction · ScanNet (TransformerFusion protocol) | F-score: 0.703 -> 0.712 | Improved |
| replace VoRTX with IOAR | 3D scene reconstruction · ScanNet (TransformerFusion protocol) | Recall: 0.651 -> 0.650 | No clear effect |
| replace VoRTX with IOAR | depth estimation from meshes · ScanNet 2D depth | Abs Rel: 0.061 -> 0.052 | Improved |
| replace VoRTX with IOAR | 3D scene reconstruction · ICL-NUIM zero-shot | F-score: 0.408 -> 0.507 | Improved |
| replace VoRTX with IOAR | 3D scene reconstruction · TUM-RGBD zero-shot | F-score: 0.229 -> 0.343 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 2 of 2 on most shared results.
- NeuralRecon — NeuralRecon: Real-Time Coherent 3D Reconstruction From Monocular Video (CVPR 2021) · better on 3 of 3 shared results
- SimpleRecon — SimpleRecon: 3D Reconstruction without 3D Convolutions (ECCV 2022) · better on 2 of 2 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
- Voxurf: Voxel-based Efficient and Accurate Neural Surface Reconstruction (ICLR 2023 · similarity 0.87)
- Fast Monocular Scene Reconstruction With Global-Sparse Local-Dense Grids (CVPR 2023 · similarity 0.86)
- UniSDF: Unifying Neural Representations for High-Fidelity 3D Reconstruction of Complex Scenes with Reflections (NeurIPS 2024 · similarity 0.85)
- GeoSVR: Taming Sparse Voxels for Geometrically Accurate Surface Reconstruction (NeurIPS 2025 · similarity 0.84)
- Fine-Detail Monocular Geometry Estimation with Self-Guided Sparse Volumetric Refinement (NeurIPS 2026 · similarity 0.84)
- Multi-view Reconstruction via SfM-guided Monocular Depth Estimation (CVPR 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.
