Research map / Point clouds / Shape and surface completion
Geometric Alignment and Prior Modulation for View-Guided Point Cloud Completion on Unseen Categories
Jingqiao Xiu, Yicong Li, Na Zhao, Han Fang, Xiang Wang, Angela Yao
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 depth lifting to 3D instead of RGB patch tokenization (ablation) | GAPM · ShapeNet-ViPC unseen | CD: 1.932 -> 1.684 | Improved |
| add depth lifting to 3D instead of RGB patch tokenization (ablation) | GAPM · ShapeNet-ViPC seen | CD: 1.166 -> 1.107 | Improved |
| add cross-source information fusion instead of plain cross-attention (ablation) | GAPM · ShapeNet-ViPC unseen | CD: 1.791 -> 1.684 | Improved |
| add attention-based region grouping instead of MLP probability grouping (ablation) | GAPM · ShapeNet-ViPC unseen | CD: 1.747 -> 1.684 | Improved |
| add learned local priors instead of modulating sampled grids (ablation) | GAPM · ShapeNet-ViPC unseen | CD: 1.889 -> 1.684 | Improved |
| add sparse prediction with region centralization (ablation) | GAPM · ShapeNet-ViPC unseen | CD: 1.821 -> 1.684 | Improved |
| use 8 local priors instead of 4 (ablation) | GAPM · ShapeNet-ViPC unseen | CD: 1.725 -> 1.684 | Improved |
| use 8 local priors instead of 16 (ablation) | GAPM · ShapeNet-ViPC unseen | CD: 1.762 -> 1.684 | Improved |
| use stronger depth estimator Depth Pro at test time without finetuning (ablation) | GAPM · ShapeNet-ViPC unseen | CD: 1.684 -> 1.643 | Improved |
| use weaker depth estimator LeReS at test time (ablation) | GAPM · ShapeNet-ViPC unseen | CD: 1.684 -> 1.813 | Hurt |
| replace EGIInet with GAPM (unseen categories) | EGIInet · ShapeNet-ViPC unseen | CD: 2.354 -> 1.684 | Improved |
| replace EGIInet with GAPM (seen categories) | EGIInet · ShapeNet-ViPC seen | CD: 1.211 -> 1.107 | Improved |
Closest papers by meaning
- Rethinking Multimodal Point Cloud Completion: A Completion-by-Correction Perspective (AAAI 2026 · similarity 0.91)
- Digging into Intrinsic Contextual Information for High-fidelity 3D Point Cloud Completion (AAAI 2025 · similarity 0.89)
- View-Guided Point Cloud Completion (CVPR 2021 · similarity 0.89)
- DC-PCN: Point Cloud Completion Network with Dual-Codebook Guided Quantization (AAAI 2025 · similarity 0.89)
- PCDreamer: Point Cloud Completion Through Multi-view Diffusion Priors (CVPR 2025 · similarity 0.88)
- Position-Aware Guided Point Cloud Completion with CLIP Model (AAAI 2025 · similarity 0.88)
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.
