Research map / 3D vision / Depth and stereo
Flow-Anything: Learning Real-World Optical Flow Estimation From Large-Scale Single-View Images
Yingping Liang, Ying Fu, Yutao Hu, Wenqi Shao, Jiaming Liu, Debing Zhang
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 dynamic objects to generated data (ablation) | Flow-Anything RAFT · KITTI 15 | EPE: 2.47 -> 1.80 | Improved |
| add dynamic objects to generated data (ablation) | Flow-Anything RAFT · KITTI 12 | EPE: 1.19 -> 1.29 | Hurt |
| add depth-aware (object-aware) inpainting (ablation) | Flow-Anything RAFT · KITTI 15 | EPE: 1.80 -> 1.85 | No clear effect |
| add depth-aware (object-aware) inpainting (ablation) | Flow-Anything RAFT · KITTI 12 | EPE: 1.29 -> 1.24 | Improved |
| add multiple moving objects (ablation) | Flow-Anything RAFT · KITTI 15 | EPE: 1.85 -> 1.79 | Improved |
| use Flow-Anything instead of MPI-Flow for data generation | RAFT · KITTI 15 (COCO images) | Fl: 10.66 -> 10.32 | Improved |
| train on FA-Flow Dataset instead of C+T (+K+S+H) | SEA-RAFT · KITTI 15 train zero-shot | EPE: 4.29 -> 1.64 | Improved |
| train on FA-Flow Dataset instead of synthetic C+T+VK2 | SEA-RAFT · KITTI 15 | EPE: 2.27 -> 1.64 | Improved |
| train on FA-Flow Dataset instead of C+T (+K+S+H) | SEA-RAFT · Sintel Final | EPE: 4.04 -> 2.22 | Improved |
| train on FA-Flow Dataset instead of C+T (+K+S+H) | FlowFormer++ · KITTI 15 train zero-shot | EPE: 3.93 -> 1.89 | Improved |
| increase generated COCO data from 20K to 120K (ablation) | RAFT from scratch · KITTI 15 | EPE: 3.68 -> 3.41 | Improved |
| use FA-Flow data instead of C+T when training from scratch | RAFT · KITTI 15 | EPE: 5.00 -> 3.41 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- MPI-Flow — MPI-Flow: Learning Realistic Optical Flow with Multiplane Images (ICCV 2023) · better on 2 of 2 shared results
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- Learning Optical Flow From Still Images (CVPR 2021 · similarity 0.93)
- SAMFlow: Eliminating Any Fragmentation in Optical Flow with Segment Anything Model (AAAI 2024 · similarity 0.90)
- Flow-NeRF: Joint Learning of Geometry, Poses, and Dense Flow within Unified Neural Representations (CVPR 2025 · similarity 0.89)
- Taming generative video models for zero-shot optical flow extraction (NeurIPS 2025 · similarity 0.89)
- DPFlow: Adaptive Optical Flow Estimation with a Dual-Pyramid Framework (CVPR 2025 · similarity 0.88)
- UnSAMFlow: Unsupervised Optical Flow Guided by Segment Anything Model (CVPR 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.
