Research map / 3D vision / Depth and stereo
Learning End-to-End Scene Flow by Distilling Single Tasks Knowledge
Filippo Aleotti, Matteo Poggi, Fabio Tosi, Stefano Mattoccia
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 |
|---|---|---|---|
| add Dense connections (ablation) | DWARF · FlyingThings3D test | Flow EPE: 7.435 -> 6.758 | Improved |
| add Dense connections (ablation) | DWARF · FlyingThings3D test | Disparity EPE: 1.959 -> 1.837 | Improved |
| add Dense connections (ablation) | DWARF · FlyingThings3D test | Change EPE: 2.283 -> 2.092 | Improved |
| add 3D correlation to the Dense variant (ablation) | DWARF · FlyingThings3D test | Flow EPE: 6.758 -> 6.738 | Improved |
| add 3D correlation to the Dense variant (ablation) | DWARF · FlyingThings3D test | Disparity EPE: 1.837 -> 1.827 | Improved |
| add 3D correlation to the Dense variant (ablation) | DWARF · FlyingThings3D test | Change EPE: 2.092 -> 2.149 | Hurt |
| add the refinement module to Dense + 3D correlation (ablation) | DWARF · FlyingThings3D test | Flow EPE: 6.738 -> 6.440 | Improved |
| add the refinement module to Dense + 3D correlation (ablation) | DWARF · FlyingThings3D test | Disparity EPE: 1.827 -> 1.784 | Improved |
| add the refinement module to Dense + 3D correlation (ablation) | DWARF · FlyingThings3D test | Change EPE: 2.149 -> 2.039 | Improved |
| use proxy labels only instead of ground-truth fine-tuning (ablation) | DWARF · KITTI 2015 validation | F1-All: 18.53 -> 20.71 | Hurt |
| use proxy labels only instead of ground-truth fine-tuning (ablation) | DWARF · KITTI 2015 validation | D1-All: 4.58 -> 3.94 | Improved |
| fine-tune first on proxy labels, then on ground truth (ablation) | DWARF · KITTI 2015 validation | F1-All: 18.53 -> 16.75 | Improved |
Closest papers by meaning
- SENSE: A Shared Encoder Network for Scene-Flow Estimation (ICCV 2019 · similarity 0.90)
- Deep Scene Flow Learning: From 2D Images to 3D Point Clouds (TPAMI 2023 · similarity 0.90)
- Flow-NeRF: Joint Learning of Geometry, Poses, and Dense Flow within Unified Neural Representations (CVPR 2025 · similarity 0.88)
- DeltaFlow: An Efficient Multi-frame Scene Flow Estimation Method (NeurIPS 2025 · similarity 0.88)
- Zero-Shot Monocular Scene Flow Estimation in the Wild (CVPR 2025 · similarity 0.87)
- Learning Temporal 3D Semantic Scene Completion via Optical Flow Guidance (NeurIPS 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.
