Research map / 3D vision / Depth and stereo
MVSTER: Epipolar Transformer for Efficient Multi-View Stereo
Xiaofeng Wang, Zheng Zhu, Guan Huang, Fangbo Qin, Yun Ye, Yijia He, Xu Chi, Xingang Wang
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 epipolar Transformer fusion instead of variance fusion (ablation) | MVSTER (D=8,8,4,4) · DTU | Overall (mm): 0.334 -> 0.313 | Improved |
| add epipolar Transformer fusion instead of variance fusion (ablation) | MVSTER (D=192) · DTU | Overall (mm): 0.460 -> 0.435 | Improved |
| add epipolar Transformer fusion instead of variance fusion (ablation) | MVSTER (D=48,32,8) · DTU | Overall (mm): 0.335 -> 0.323 | Improved |
| use epipolar Transformer instead of CNN fusion (ablation) | MVSTER (D=8,8,4,4) · DTU | Runtime (s): 0.17 -> 0.09 | Improved |
| use epipolar Transformer instead of CNN fusion (ablation) | MVSTER (D=8,8,4,4) · DTU | Overall (mm): 0.320 -> 0.313 | Improved |
| add monocular depth estimator auxiliary branch (ablation) | MVSTER · DTU | Overall (mm): 0.317 -> 0.313 | Improved |
| use MDE instead of DCN (ablation) | MVSTER · DTU | Runtime (s): 0.23 -> 0.09 | Improved |
| use MDE instead of ASPP (ablation) | MVSTER · DTU | Overall (mm): 0.327 -> 0.313 | Improved |
| use optimal transport (Wasserstein) loss instead of L1 loss (ablation) | MVSTER · DTU | Overall (mm): 0.321 -> 0.313 | Improved |
| use optimal transport loss instead of cross-entropy loss (ablation) | MVSTER · DTU | Stage 4 EPE: 6.64 -> 5.90 | Improved |
| lower loss weight lambda from 1x10^-2 to 3x10^-4 (ablation) | MVSTER · DTU | Overall (mm): 0.337 -> 0.313 | Improved |
| raise temperature t_e from 0.5 to 2.0 (ablation) | MVSTER · DTU | Overall (mm): 0.317 -> 0.313 | Improved |
Later papers that compared with this method
4 later accepted papers report MVSTER as a baseline in their results tables; 4 of them report a better number on most of the results they share with it.
- RRT-MVS: Recurrent Regularization Transformer for Multi-View Stereo (AAAI 2025 · better on 1 of 1 shared result)
- Hierarchical Prior Mining for Non-local Multi-View Stereo (ICCV 2023 · better on 3 of 3 shared results)
- When Epipolar Constraint Meets Non-Local Operators in Multi-View Stereo (ICCV 2023 · better on 3 of 3 shared results)
- GeoMVSNet: Learning Multi-View Stereo With Geometry Perception (CVPR 2023 · better on 1 of 1 shared result)
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 4 of 5 on most shared results.
- IterMVS — IterMVS: Iterative Probability Estimation for Efficient Multi-View Stereo (CVPR 2022) · better on 1 of 2 shared results
- TransMVSNet — TransMVSNet: Global Context-Aware Multi-View Stereo Network With Transformers (CVPR 2022) · better on 2 of 2 shared results
- PatchmatchNet — PatchmatchNet: Learned Multi-View Patchmatch Stereo (CVPR 2021) · better on 1 of 1 shared result
- AA-RMVSNet — AA-RMVSNet: Adaptive Aggregation Recurrent Multi-View Stereo Network (ICCV 2021) · better on 1 of 1 shared result
- EPP-MVSNet — EPP-MVSNet: Epipolar-Assembling Based Depth Prediction for Multi-View Stereo (ICCV 2021) · better on 1 of 1 shared result
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
- When Epipolar Constraint Meets Non-Local Operators in Multi-View Stereo (ICCV 2023 · similarity 0.91)
- MVSFormer++: Revealing the Devil in Transformer's Details for Multi-View Stereo (ICLR 2024 · similarity 0.90)
- EPP-MVSNet: Epipolar-Assembling Based Depth Prediction for Multi-View Stereo (ICCV 2021 · similarity 0.89)
- WT-MVSNet: Window-based Transformers for Multi-view Stereo (NeurIPS 2022 · similarity 0.89)
- A Global Depth-Range-Free Multi-View Stereo Transformer Network with Pose Embedding (NeurIPS 2024 · similarity 0.89)
- RRT-MVS: Recurrent Regularization Transformer for Multi-View Stereo (AAAI 2025 · similarity 0.87)
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.
