Research map / Video / Video segmentation and tracking
EMatch: A Unified Framework for Event-based Optical Flow and Stereo Matching
Pengjie Zhang, Lin Zhu, Xiao Wang, Lizhi Wang, Hua Huang
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 |
|---|---|---|---|
| train a single model on both flow and stereo (multi-task) instead of task-specific models | EMatch · DSEC optical flow | EPE: 0.67 -> 0.64 | Improved |
| initialize from the shared module of the other task (cross-task transfer) | EMatch · DSEC optical flow | EPE: 0.67 -> 0.65 | Improved |
| initialize from the shared module of the other task (cross-task transfer) | EMatch · DSEC stereo | 1PE: 8.80 -> 8.44 | Improved |
| add Temporal Recurrent Network instead of CNN pyramid (ablation) | EMatch · DSEC flow (scaled-down) | EPE: 1.05 -> 0.94 | Improved |
| add convGRU recurrence in TRN (ablation) | EMatch · DSEC flow (scaled-down) | EPE: 0.99 -> 0.94 | Improved |
| add Spatial Contextual Attention (ablation) | EMatch · DSEC flow (scaled-down) | EPE: 1.20 -> 0.94 | Improved |
| add attention operations in SCA (ablation) | EMatch · DSEC flow (scaled-down) | 1PE: 25.80 -> 15.38 | Improved |
| add multi-scale optimization (ablation) | EMatch · DSEC flow (scaled-down) | EPE: 1.27 -> 0.94 | Improved |
| add task-specific refinement heads (ablation) | EMatch · DSEC flow (scaled-down) | EPE: 1.56 -> 0.94 | Improved |
| train unified model on both flow and disparity data | EMatch · sparser event input (generalization) | degradation under reduced dt or event count | Improved |
Closest papers by meaning
- Temporal Event Stereo via Joint Learning with Stereoscopic Flow (ECCV 2024 · similarity 0.93)
- Temporal Stereo Matching From Event Cameras via Joint Learning With Stereoscopic Flow (TPAMI 2025 · similarity 0.93)
- Bridging Stereo Matching and Optical Flow via Spatiotemporal Correspondence (CVPR 2019 · similarity 0.91)
- EDCFlow: Exploring Temporally Dense Difference Maps for Event-based Optical Flow Estimation (CVPR 2025 · similarity 0.89)
- Learning an Event Sequence Embedding for Dense Event-Based Deep Stereo (ICCV 2019 · similarity 0.89)
- ARES: Unifying Asymmetric RGB-Event Stereo for Probabilistic Scene Flow Estimation (CVPR 2026 · similarity 0.89)
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.
