Research map / 3D vision / Depth and stereo
Global Regulation and Excitation via Attention Tuning for Stereo Matching
Jiahao Li, Xinhong Chen, Zhengmin Jiang, Qian Zhou, Yung-Hui Li, Jianping 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 Volume Attention with a correlation cost volume (ablation) | IGEV-Stereo · Scene Flow test | EPE all (px): 0.479 -> 0.469 | Improved |
| replace correlation cost volume with concatenation cost volume under Volume Attention (ablation) | IGEV-Stereo + VA · Scene Flow test | EPE all (px): 0.469 -> 0.466 | Improved |
| add Spatial Attention to concatenation cost volume with Volume Attention (ablation) | IGEV-Stereo + VA · Scene Flow test | EPE occluded (px): 1.616 -> 1.587 | Improved |
| add Matching Attention (without VA or SA) to concatenation cost volume (ablation) | IGEV-Stereo · Scene Flow test | EPE non-occluded (px): 0.194 -> 0.153 | Improved |
| add SA and MA to Volume Attention (full model) (ablation) | GREAT-IGEV · Scene Flow test | EPE all (px): 0.466 -> 0.413 | Improved |
| integrate the GREAT attention modules into RAFT-Stereo | RAFT-Stereo · Scene Flow test | EPE all (px): 0.551 -> 0.488 | Improved |
| integrate the GREAT attention modules into IGEV-Stereo | IGEV-Stereo · Scene Flow test | EPE all (px): 0.479 -> 0.413 | Improved |
| integrate the GREAT attention modules into Selective-IGEV | Selective-IGEV · Scene Flow test | EPE all (px): 0.453 -> 0.419 | Improved |
| integrate the GREAT modules into IGEV-Stereo, occluded regions | IGEV-Stereo · Scene Flow test | EPE occluded (px): 1.649 -> 1.514 | Improved |
| integrate the GREAT modules into IGEV-Stereo, non-occluded regions | IGEV-Stereo · Scene Flow test | EPE non-occluded (px): 0.194 -> 0.135 | Improved |
| use GREAT attention modules with fewer iterations | GREAT-IGEV · Scene Flow test | EPE all (px): 0.483 at 4 iterations vs 0.479 for IGEV-Stereo at 32 | Improved |
| integrate GREAT modules into RAFT-Stereo, zero-shot to KITTI 2015 | RAFT-Stereo · KITTI 2015 (zero-shot) | 3-pixel error (%): 5.8 -> 5.8 | No clear effect |
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.
- IGEV++ — IGEV++: Iterative Multi-Range Geometry Encoding Volumes for Stereo Matching (TPAMI 2025) · 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
- Accurate and Efficient Stereo Matching via Attention Concatenation Volume (TPAMI 2023 · similarity 0.88)
- Selective-Stereo: Adaptive Frequency Information Selection for Stereo Matching (CVPR 2024 · similarity 0.88)
- Stereo Anywhere: Robust Zero-Shot Deep Stereo Matching Even Where Either Stereo or Mono Fail (CVPR 2025 · similarity 0.86)
- Any-Stereo: Arbitrary Scale Disparity Estimation for Iterative Stereo Matching (AAAI 2024 · similarity 0.85)
- Mono2Stereo: A Benchmark and Empirical Study for Stereo Conversion (CVPR 2025 · similarity 0.85)
- TurboVGGT: Fast Visual Geometry Reconstruction with Adaptive Alternating Attention (NeurIPS 2026 · 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.
