Research map / 3D vision / Depth and stereo
GyroFlow: Gyroscope-Guided Unsupervised Optical Flow Learning
Haipeng Li, Kunming Luo, Shuaicheng Liu
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 self-guided fusion (SGF) module (ablation) | GyroFlow · GOF evaluation set (average) | EPE: 0.95 -> 0.72 | Improved |
| add SGF module (ablation) | UnFlow · GOF evaluation set (average) | EPE: 1.22 -> 0.90 | Improved |
| add SGF module (ablation) | ARFlow · GOF evaluation set (average) | EPE: 1.04 -> 0.86 | Improved |
| add SGF module (ablation) | UFlow · GOF evaluation set (average) | EPE: 0.98 -> 0.80 | Improved |
| use full SGF instead of fusion-flow-only variant (SGF-Fuse) (ablation) | GyroFlow · GOF evaluation set (average) | EPE: 0.80 -> 0.72 | Improved |
| use full SGF instead of map-only variant (SGF-Map) (ablation) | GyroFlow · GOF evaluation set (average) | EPE: 0.90 -> 0.72 | Improved |
| use separate fusion and map blocks instead of one dense block (ablation) | GyroFlow · GOF evaluation set (average) | EPE: 0.89 -> 0.72 | Improved |
| use SGF instead of directly adding gyro field to flow (DPGF) (ablation) | GyroFlow · GOF evaluation set (average) | EPE: 0.98 -> 0.72 | Improved |
| use SGF instead of warping image with gyro field (DWI) (ablation) | GyroFlow · GOF evaluation set (average) | EPE: 4.38 -> 0.72 | Improved |
| fuse gyro field at all pyramid layers instead of only 1/32 resolution (ablation) | GyroFlow · GOF evaluation set (average) | EPE: 0.95 -> 0.72 | Improved |
| fuse gyro field at 1/4 resolution instead of 1/32 resolution (ablation) | GyroFlow · GOF evaluation set (average) | EPE: 0.95 -> 0.87 | Improved |
Closest papers by meaning
- GyroFlow+: Gyroscope-Guided Unsupervised Deep Homography and Optical Flow Learning (IJCV 2024 · similarity 0.96)
- Gyro-based Deep Video Deblurring (CVPR 2026 · similarity 0.86)
- Gyro-based Neural Single Image Deblurring (CVPR 2025 · similarity 0.86)
- Flow-NeRF: Joint Learning of Geometry, Poses, and Dense Flow within Unified Neural Representations (CVPR 2025 · similarity 0.84)
- The Right Spin: Learning Object Motion from Rotation-Compensated Flow Fields (IJCV 2023 · similarity 0.83)
- MoE-Gyro: Self-Supervised Over-Range Reconstruction and Denoising for MEMS Gyroscopes (NeurIPS 2025 · similarity 0.82)
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.
