Research map / 3D vision / Depth and stereo
Just a Few Points Are All You Need for Multi-View Stereo: A Novel Semi-Supervised Learning Method for Multi-View Stereo
Taekyung Kim, Jaehoon Choi, Seokeon Choi, Dongki Jung, Changick Kim
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 double prediction layers (ablation) | SGT-MVSNet baseline · DTU eval (sparse 1e-5) | overall (mm): 0.441 -> 0.434 | Improved |
| add double prediction layers (ablation) | SGT-MVSNet baseline · DTU eval (sparse 1e-5) | accuracy (mm): 0.460 -> 0.448 | Improved |
| add 3D Point Consistency Loss (ablation) | SGT-MVSNet baseline · DTU eval (sparse 1e-5) | overall (mm): 0.434 -> 0.417 | Improved |
| add 3D Point Consistency Loss (ablation) | SGT-MVSNet baseline · DTU eval (sparse 1e-5) | accuracy (mm): 0.448 -> 0.438 | Improved |
| add 3D Point Consistency Loss (ablation) | SGT-MVSNet baseline · DTU eval (sparse 1e-5) | completeness (mm): 0.419 -> 0.396 | Improved |
| add Coarse-to-fine Reliable Depth Propagation (ablation) | SGT-MVSNet baseline · DTU eval (sparse 1e-5) | completeness (mm): 0.396 -> 0.381 | Improved |
| add Coarse-to-fine Reliable Depth Propagation (ablation) | SGT-MVSNet baseline · DTU eval (sparse 1e-5) | accuracy (mm): 0.438 -> 0.441 | Hurt |
| add Coarse-to-fine Reliable Depth Propagation (ablation) | SGT-MVSNet baseline · DTU eval (sparse 1e-5) | overall (mm): 0.417 -> 0.411 | Improved |
| use SGT-MVSNet instead of Fast-MVSNet | sparse-GT MVS training · DTU eval (sparse 1e-5) | overall (mm): 0.431 -> 0.411 | Improved |
| use SGT-MVSNet instead of Fast-MVSNet | sparse-GT MVS training · DTU eval (sparse 1e-5) | accuracy (mm): 0.381 -> 0.441 | Hurt |
| increase sparse ground truth density from 1e-5 to 1e-4 | SGT-MVSNet · DTU eval | overall (mm): 0.411 -> 0.385 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- Fast-MVSNet — Fast-MVSNet: Sparse-to-Dense Multi-View Stereo With Learned Propagation and Gauss-Newton Refinement (CVPR 2020) · better on 4 of 6 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
- Fast-MVSNet: Sparse-to-Dense Multi-View Stereo With Learned Propagation and Gauss-Newton Refinement (CVPR 2020 · similarity 0.93)
- Learning-Based Multi-View Stereo: A Survey (TPAMI 2026 · similarity 0.92)
- Point-Based Multi-View Stereo Network (ICCV 2019 · similarity 0.92)
- Generalized Binary Search Network for Highly-Efficient Multi-View Stereo (CVPR 2022 · similarity 0.92)
- MVS2D: Efficient Multi-View Stereo via Attention-Driven 2D Convolutions (CVPR 2022 · similarity 0.91)
- P-MVSNet: Learning Patch-Wise Matching Confidence Aggregation for Multi-View Stereo (ICCV 2019 · similarity 0.91)
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.
