Research map / Video / Video segmentation and tracking
Match-Stereo-Videos: Bidirectional Alignment for Consistent Dynamic Stereo Matching
Junpeng Jing, Ye Mao, Krystian Mikolajczyk
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 |
|---|---|---|---|
| replace 2D conv updater with 3D conv updater (ablation) | per-frame baseline · Sintel final | TEPE: 1.74 -> 1.34 | Improved |
| replace 2D conv with 3D conv updater (ablation) | per-frame baseline · Dynamic Replica | delta_1px (accuracy): 2.40 -> 3.38 | Hurt |
| enlarge updater kernel to 1x1x15 horizontally (ablation) | 3D conv updater · Dynamic Replica | TEPE: 0.082 -> 0.079 | Improved |
| use separated instead of shared weights for update modules (ablation) | BiDAStereo updater · Sintel final | TEPE: 1.33 -> 1.34 | No clear effect |
| add bidirectional alignment with single left frame in correlation (ablation) | BiDAStereo · Dynamic Replica | TEPE: 0.079 -> 0.067 | Improved |
| use multiple left reference frames (multi-multi) instead of single left frame (ablation) | triple-frame correlation · Dynamic Replica | TEPE: 0.067 -> 0.082 | Hurt |
| share motion hidden state across scales instead of re-initializing (ablation) | motion-propagation recurrent unit · Dynamic Replica | TEPE: 0.511 -> 0.062 | Improved |
| increase inference frames from 10 to 20 (ablation) | BiDAStereo · Dynamic Replica | TEPE: 0.081 -> 0.062 | Improved |
| increase inference iterations from 20 to 32 (ablation) | BiDAStereo · Sintel final | TEPE: 1.24 -> 1.23 | No clear effect |
| add Dynamic Replica to SceneFlow training data | BiDAStereo · Sintel final | TEPE: 1.26 -> 1.22 | Improved |
| use BiDAStereo instead of CREStereo on Sintel clean | stereo matching · Sintel clean | TEPE: 0.67 -> 0.75 | Hurt |
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.
- DynamicStereo — DynamicStereo: Consistent Dynamic Depth From Stereo Videos (CVPR 2023) · better on 5 of 5 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
- Stereo Any Video: Temporally Consistent Stereo Matching (ICCV 2025 · similarity 0.92)
- Match Stereo Videos via Bidirectional Alignment (TPAMI 2026 · similarity 0.92)
- PPMStereo: Pick-and-Play Memory Construction for Consistent Dynamic Stereo Matching (NeurIPS 2025 · similarity 0.92)
- Temporally Consistent Stereo Matching (ECCV 2024 · similarity 0.90)
- DynamicStereo: Consistent Dynamic Depth From Stereo Videos (CVPR 2023 · similarity 0.90)
- Any-Stereo: Arbitrary Scale Disparity Estimation for Iterative Stereo Matching (AAAI 2024 · 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.
