Research map / 3D vision / Depth and stereo
RobuSTereo: Robust Zero-Shot Stereo Matching under Adverse Weather
Yuran Wang, Yingping Liang, Yutao Hu, Ying Fu
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 coherence-enhanced consistency module to data generation (ablation) | StereoBase on RST-Dataset · DrivingStereo | EPE and D1: 1.308 -> 0.875 EPE; 5.997 -> 2.050 D1 | Improved |
| use KITTI instead of vKITTI as source images for generation (ablation) | StereoBase on RST-Dataset · DrivingStereo | EPE and D1: 1.039 -> 0.875 EPE; 4.689 -> 2.050 D1 | Improved |
| use robust encoder instead of MobileNetV2 (ablation) | StereoBase · DrivingStereo | EPE and D1: 0.875 -> 0.836 EPE; 2.050 -> 1.598 D1 | Improved |
| use robust encoder instead of Dinov2 (ablation) | StereoBase · DrivingStereo | EPE and D1: 0.852 -> 0.836 EPE; 1.793 -> 1.598 D1 | Improved |
| train PSMNet on RST-Dataset instead of SceneFlow | PSMNet · DrivingStereo | overall EPE: 12.133 -> 1.064 | Improved |
| train IGEV on RST-Dataset instead of SceneFlow | IGEV · DrivingStereo | overall EPE: 1.356 -> 0.943 | Improved |
| train StereoBase on RST-Dataset instead of SceneFlow | StereoBase · DrivingStereo | overall EPE: 1.302 -> 0.875 | Improved |
| train StereoBase on RST-Dataset instead of KITTI | StereoBase · DrivingStereo | overall EPE: 0.927 -> 0.875 | Improved |
| train IGEV on RST-Dataset instead of KITTI | IGEV · DrivingStereo | overall EPE and sunny D1: 1.009 -> 0.943 overall EPE; sunny D1 2.026 -> 2.308 | Mixed |
| train StereoBase on RST-Dataset instead of KITTI | StereoBase · SeeingThroughFog | overall EPE: 6.976 -> 6.158 EPE; D1 26.449 -> 24.026 | Improved |
| add robust encoder and RST-Dataset training over StereoBase on RST-Dataset | StereoBase · SeeingThroughFog | overall EPE: 6.158 -> 3.359 | Improved |
| train RobuSTereo on RST-Dataset vs mixed-data zero-shot models | RobuSTereo · DrivingStereo | overall EPE: 1.042 -> 0.836 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 3 of 3 on most shared results.
- MonSter — MonSter: Marry Monodepth to Stereo Unleashes Power (CVPR 2025) · better on 2 of 2 shared results
- IGEV — IGEV++: Iterative Multi-Range Geometry Encoding Volumes for Stereo Matching (TPAMI 2025) · better on 1 of 1 shared result
- PSMNet — PSMNet: Position-Aware Stereo Merging Network for Room Layout Estimation (CVPR 2022) · 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
- Weather-Conditioned Depth Anything (ECCV 2026 · similarity 0.88)
- Any-Stereo: Arbitrary Scale Disparity Estimation for Iterative Stereo Matching (AAAI 2024 · similarity 0.86)
- Synthetic-to-Real Self-supervised Robust Depth Estimation via Learning with Motion and Structure Priors (CVPR 2025 · similarity 0.85)
- FoundationStereo: Zero-Shot Stereo Matching (CVPR 2025 · similarity 0.85)
- Stereo Anywhere: Robust Zero-Shot Deep Stereo Matching Even Where Either Stereo or Mono Fail (CVPR 2025 · similarity 0.85)
- Robust Synthetic-to-Real Transfer for Stereo Matching (CVPR 2024 · similarity 0.85)
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.
