Research map / Architectures and efficiency / RNNs and normalization
Practical Stereo Matching via Cascaded Recurrent Network With Adaptive Correlation
Check your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 adaptive group correlation with deformable search, group-wise correlation, and attention (ablation) | CREStereo · Middlebury 2014 | Bad 2.0: 6.86 -> 6.46 | Helped |
| add deformable search (ablation) | CREStereo · Middlebury 2014 | Bad 2.0: 6.84 -> 6.46 | Helped |
| add group-wise correlation (ablation) | CREStereo · Middlebury 2014 | Bad 2.0: 6.82 -> 6.46 | Helped |
| add local feature attention (ablation) | CREStereo · Middlebury 2014 | Bad 2.0: 6.49 -> 6.46 | Helped |
| use alternating 1D and 2D local correlation instead of only 1D or only 2D local correlation (ablation) | CREStereo · Middlebury 2014 | Bad 2.0: 19.87 -> 19.23; 20.70 -> 19.23 | Helped |
| replace local correlation with all-pairs correlation (ablation) | CREStereo · Middlebury 2014 | Bad 2.0: 19.87 -> 47.38; 19.87 -> 44.41 | Hurt |
| use 2 cascade levels rather than 1D+2D local correlation without multiple levels (ablation) | CREStereo · Middlebury 2014 | Bad 2.0: 19.23 -> 12.48 | Helped |
| increase from 2 to 3 cascade levels (ablation) | CREStereo · Middlebury 2014 | AvgErr: 1.99 -> 1.80 | Helped |
| use 2 stacked cascades instead of a single cascade at 1024 × 1536 (ablation) | CREStereo · Middlebury 2014 | Bad 2.0: 6.00 -> 5.30 | Helped |
| increase stacked cascades from 2 to 3 at 1536 × 2048 (ablation) | CREStereo · Middlebury 2014 | Bad 2.0: 4.53 -> 4.58 | No clear effect |
| train on the proposed synthetic dataset rather than a similarly sized SceneFlow sample | CREStereo · ETH3D and Middlebury validation | validation error | Helped |
| use CREStereo rather than RAFT-Stereo | stereo matching methods · Middlebury 2014 test | Bad 2.0: 4.74 -> 3.71 | Helped |
Closest papers by meaning
- Local Similarity Pattern and Cost Self-Reassembling for Deep Stereo Matching Networks (AAAI 2022 · similarity 0.91)
- Zoom and Learn: Generalizing Deep Stereo Matching to Novel Domains (CVPR 2018 · similarity 0.90)
- AANet: Adaptive Aggregation Network for Efficient Stereo Matching (CVPR 2020 · similarity 0.90)
- Stereo Risk: A Continuous Modeling Approach to Stereo Matching (ICML 2024 · similarity 0.89)
- Hierarchical Object-Aware Dual-Level Contrastive Learning for Domain Generalized Stereo Matching (NeurIPS 2024 · similarity 0.88)
- Rectification-specific Supervision and Constrained Estimator for Online Stereo Rectification (CVPR 2025 · similarity 0.88)
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.