Research map / 3D vision / Depth and stereo
SMD-Nets: Stereo Mixture Density Networks
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 the bimodal SMD-Net output representation to PSM | PSM · UnrealStereo4K in-domain | SEE3 Avg: 1.73 -> 1.13 | Helped |
| Add the bimodal SMD-Net output representation to PSM | PSM · KITTI 2015 validation boundaries | SEE3 Avg: 1.10 -> 0.90 | Helped |
| Add the bimodal SMD-Net output representation to PSM (ablation) | PSM · UnrealStereo4K binocular stereo | SEE3 Avg: 1.98 -> 1.52 | Helped |
| Replace a point estimate with a unimodal Laplacian output (ablation) | PSM · UnrealStereo4K binocular stereo | SEE3 Avg: 1.98 -> 2.50 | Hurt |
| Use a bimodal output instead of a unimodal Laplacian (ablation) | PSM · UnrealStereo4K binocular stereo | SEE3 Avg: 2.50 -> 1.52 | Helped |
| Replace random sampling with DDA sampling at rho 10 (ablation) | PSM + SMD-Net · UnrealStereo4K test | SEE3 Avg: 1.52 -> 1.13 | Helped |
| Use DDA sampling at rho 10 instead of rho 0 (ablation) | PSM + SMD-Net · UnrealStereo4K test | SEE3 Avg: 1.34 -> 1.13 | Helped |
| Increase DDA dilation from rho 10 to rho 20 (ablation) | PSM + SMD-Net · UnrealStereo4K test | SEE3 Avg: 1.13 -> 1.30 | Hurt |
| Use DDA sampling at rho 10 instead of random sampling (ablation) | PSM + SMD-Net · UnrealStereo4K test | EPE Avg: 1.11 -> 0.92 | Helped |
| Replace the point estimate with a bimodal output (ablation) | U-Net, PSM, HSM · UnrealStereo4K binocular, monocular, and active depth | SEE3 Avg | Helped |
Closest papers by meaning
- CFNet: Cascade and Fused Cost Volume for Robust Stereo Matching (CVPR 2021 · similarity 0.90)
- AANet: Adaptive Aggregation Network for Efficient Stereo Matching (CVPR 2020 · similarity 0.89)
- Guided Stereo Matching (CVPR 2019 · similarity 0.89)
- Cost Volume Pyramid Based Depth Inference for Multi-View Stereo (CVPR 2020 · similarity 0.89)
- Stereo Risk: A Continuous Modeling Approach to Stereo Matching (ICML 2024 · similarity 0.88)
- Any-Stereo: Arbitrary Scale Disparity Estimation for Iterative Stereo Matching (AAAI 2024 · 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.