Research map / Label-efficient and robust learning / Test-time adaptation
Real-Time Self-Adaptive Deep Stereo
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 |
|---|---|---|---|
| Apply unsupervised full online adaptation | MADNet · KITTI City | D1-all (%): 37.42 -> 2.63 | Helped |
| Apply unsupervised full online adaptation | MADNet · KITTI Residential | D1-all (%): 37.41 -> 2.44 | Helped |
| Apply unsupervised full online adaptation | MADNet · KITTI Campus | D1-all (%): 51.98 -> 8.91 | Helped |
| Apply unsupervised full online adaptation | MADNet · KITTI Road | D1-all (%): 47.45 -> 2.33 | Helped |
| Use modular online adaptation instead of no adaptation | MADNet · KITTI raw, concatenated environments | D1-all (%): 38.84 -> 3.37 | Helped |
| Use modular online adaptation rather than full back-propagation | MADNet · KITTI raw, concatenated environments | FPS: 14.26 -> 25.43 | Helped |
| Adapt with MADNet rather than DispNetC | Full online adaptation · KITTI City | D1-all (%): 4.34 -> 2.63 | Helped |
| Adapt with MADNet rather than DispNetC | Full online adaptation · KITTI Campus | D1-all (%): 8.66 -> 8.91 | Hurt |
| Use online full adaptation rather than no adaptation | DispNetC · KITTI Residential | D1-all (%): 8.72 -> 3.60 | Helped |
| Adapt only the last layer instead of modularly training across the network (ablation) | MADNet · KITTI raw, concatenated environments | D1-all (%): 3.37 -> 38.33 | Hurt |
| Adapt only the Refinement module instead of modularly training across the network (ablation) | MADNet · KITTI raw, concatenated environments | D1-all (%): 3.37 -> 31.89 | Hurt |
| Adapt only D2 and Refinement modules instead of modularly training across the network (ablation) | MADNet · KITTI raw, concatenated environments | D1-all (%): 3.37 -> 18.84 | Hurt |
Closest papers by meaning
- Learning to Adapt for Stereo (CVPR 2019 · similarity 0.89)
- SMD-Nets: Stereo Mixture Density Networks (CVPR 2021 · similarity 0.88)
- AANet: Adaptive Aggregation Network for Efficient Stereo Matching (CVPR 2020 · similarity 0.88)
- Any-Stereo: Arbitrary Scale Disparity Estimation for Iterative Stereo Matching (AAAI 2024 · similarity 0.86)
- IINet: Implicit Intra-inter Information Fusion for Real-Time Stereo Matching (AAAI 2024 · similarity 0.86)
- Stereo Risk: A Continuous Modeling Approach to Stereo Matching (ICML 2024 · similarity 0.86)
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.