Research map / Probabilistic and scientific modeling / Molecules and proteins
DDFlow: Learning Optical Flow with Unlabeled Data Distillation
Read the paperCheck 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 |
|---|---|---|---|
| Distill teacher flow predictions to supervise cropped-patch occlusions (ablation) | DDFlow · Flying Chairs test | EPE: 3.24 -> 2.97 | Helped |
| Add occlusion handling (ablation) | DDFlow without census or distillation · Flying Chairs test | EPE: 4.06 -> 3.95 | Helped |
| Add occlusion handling (ablation) | DDFlow with census transform and without distillation · Flying Chairs test | EPE: 3.75 -> 3.24 | Helped |
| Add census transform (ablation) | DDFlow without occlusion handling or distillation · Flying Chairs test | EPE: 4.06 -> 3.75 | Helped |
| Add census transform (ablation) | DDFlow with occlusion handling and without distillation · Flying Chairs test | EPE: 3.95 -> 3.24 | Helped |
| Add data distillation (ablation) | DDFlow with occlusion handling and census transform · Sintel Clean train | EPE OCC: 29.36 -> 23.92 | Helped |
| Add data distillation (ablation) | DDFlow with occlusion handling and census transform · Sintel Final train | EPE OCC: 31.86 -> 26.74 | Helped |
| Add data distillation (ablation) | DDFlow with occlusion handling and census transform · KITTI 2012 train | EPE OCC: 27.04 -> 11.31 | Helped |
| Add data distillation (ablation) | DDFlow with occlusion handling and census transform · KITTI 2015 train | EPE OCC: 42.66 -> 24.68 | Helped |
| Fine-tune DDFlow on Sintel training data | DDFlow pretrained on Flying Chairs · Sintel Clean test | EPE: 7.23 -> 6.18 | Helped |
| Fine-tune DDFlow on Sintel training data | DDFlow pretrained on Flying Chairs · Sintel Final test | EPE: 8.81 -> 7.40 | Helped |
| Use DDFlow instead of the strongest reported unsupervised comparator | Optical flow estimation · Flying Chairs test | EPE: 3.30 -> 2.97 | Helped |
Closest papers by meaning
- UnFlow: Unsupervised Learning of Optical Flow With a Bidirectional Census Loss (AAAI 2018 · similarity 0.91)
- DFlow: Learning to Synthesize Better Optical Flow Datasets via a Differentiable Pipeline (ICLR 2023 · similarity 0.91)
- OCAI: Improving Optical Flow Estimation by Occlusion and Consistency Aware Interpolation (CVPR 2024 · similarity 0.89)
- UnSAMFlow: Unsupervised Optical Flow Guided by Segment Anything Model (CVPR 2024 · similarity 0.89)
- DPFlow: Adaptive Optical Flow Estimation with a Dual-Pyramid Framework (CVPR 2025 · similarity 0.88)
- Recurrent Partial Kernel Network for Efficient Optical Flow Estimation (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.