Research map / 3D vision / Depth and stereo
Optical Flow in Mostly Rigid Scenes
Jonas Wulff, Laura Sevilla-Lara, Michael J. Black
Check 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 |
|---|---|---|---|
| Use MR-Flow instead of DF initialization | Optical flow · MPI-Sintel Clean test | EPE: 3.57 -> 2.53 | Improved |
| Use MR-Flow instead of DF initialization | Optical flow · MPI-Sintel Final test | EPE: 6.08 -> 5.38 | Improved |
| Use MR-Flow instead of SDF | Optical flow · KITTI 2015 test | incorrect pixels: 11.01% -> 12.19% | Hurt |
| Add occlusion reasoning (ablation) | MR-Flow · MPI-Sintel Final reduced train, all pixels | EPE: 3.759 -> 3.614 | Improved |
| Add coplanarity refinement (ablation) | MR-Flow · MPI-Sintel Final reduced train, all pixels | EPE: 3.645 -> 3.614 | Improved |
| Add nonlinear initialization of b− (ablation) | MR-Flow · MPI-Sintel Final reduced train, all pixels | EPE: 3.656 -> 3.614 | Improved |
| Add spatial priors (ablation) | MR-Flow · MPI-Sintel Final reduced train, all pixels | EPE: 3.628 -> 3.614 | Improved |
| Add variational refinement using only the data term (ablation) | MR-Flow · MPI-Sintel Final reduced train, all pixels | EPE: 3.6214 -> 3.6285 | Hurt |
| Add second-order regularization (ablation) | MR-Flow variational refinement · MPI-Sintel Final reduced train, all pixels | EPE: 3.6274 -> 3.6138 | Improved |
| Add first-order regularization (ablation) | MR-Flow variational refinement · MPI-Sintel Final reduced train, all pixels | EPE: 3.6138 -> 3.6138 | No clear effect |
Closest papers by meaning
- Learning Rigidity in Dynamic Scenes with a Moving Camera for 3D Motion Field Estimation (ECCV 2018 · similarity 0.89)
- Optical Flow With Semantic Segmentation and Localized Layers (CVPR 2016 · similarity 0.89)
- Fast Multi-Frame Stereo Scene Flow With Motion Segmentation (CVPR 2017 · similarity 0.89)
- MPI-Flow: Learning Realistic Optical Flow with Multiplane Images (ICCV 2023 · similarity 0.88)
- Learning To Segment Rigid Motions From Two Frames (CVPR 2021 · similarity 0.88)
- Zero-Shot Monocular Scene Flow Estimation in the Wild (CVPR 2025 · 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.
