Research map / Point clouds / Shape and surface completion
Point-Based Multi-View Stereo Network
Rui Chen, Songfang Han, Jing Xu, Hao Su
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 edge convolution (relative-feature aggregation) (ablation) | PointFlow · DTU | accuracy / completeness (mm): accuracy 0.501 -> 0.448; completeness 0.518 -> 0.487 | Improved |
| add Euclidean kNN grouping instead of pixel neighbours (ablation) | PointFlow · DTU | accuracy / completeness (mm): accuracy 0.455 -> 0.448; completeness 0.492 -> 0.487 | Improved |
| add image feature pyramid (ablation) | PointFlow · DTU | accuracy / completeness (mm): accuracy 0.455 -> 0.448; completeness 0.489 -> 0.487 | Improved |
| more flow iterations (0 to 3) (ablation) | Point-MVSNet · DTU | overall (mm): 0.726 -> 0.391 | Improved |
| one flow iteration without upsampling vs coarse (ablation) | Point-MVSNet · DTU | overall (mm): 0.726 -> 0.712 | Improved |
| increase point hypotheses from m=2 to m=1 (ablation) | Point-MVSNet · DTU | overall (mm): 0.468 -> 0.479 | Hurt |
| increase point hypotheses from m=2 to m=3 (ablation) | Point-MVSNet · DTU | overall (mm): 0.468 -> 0.484 | Hurt |
| more input views (2 to 5) (ablation) | Point-MVSNet · DTU | overall (mm): 0.533 -> 0.468 | Improved |
| replace PU-Net upsampling with PointFlow | coarse point cloud · DTU | overall (mm): 0.943 -> 0.391 | Improved |
| apply PointFlow refinement to MVSNet depth | MVSNet · Tanks and Temples intermediate | f-score: 43.48 -> 48.27 | Improved |
| use pixel neighbours instead of Euclidean kNN (ablation) | Point-MVSNet iteration 2 · DTU | overall (mm): 0.468 -> 0.474 | No clear effect |
| max pooling vs average pooling in edge convolution (ablation) | PointFlow · DTU | reconstruction error | No clear effect |
Closest papers by meaning
- Visibility-Aware Point-Based Multi-View Stereo Network (TPAMI 2020 · similarity 0.96)
- Just a Few Points Are All You Need for Multi-View Stereo: A Novel Semi-Supervised Learning Method for Multi-View Stereo (ICCV 2021 · similarity 0.92)
- Pyramid Multi-view Stereo Net with Self-adaptive View Aggregation (ECCV 2020 · similarity 0.91)
- Fast-MVSNet: Sparse-to-Dense Multi-View Stereo With Learned Propagation and Gauss-Newton Refinement (CVPR 2020 · similarity 0.91)
- MVSNet: Depth Inference for Unstructured Multi-view Stereo (ECCV 2018 · similarity 0.91)
- Multi-View Stereo Representation Revist: Region-Aware MVSNet (CVPR 2023 · similarity 0.90)
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.
