Research map / Video / Video segmentation and tracking
Event-based Visual Deformation Measurement
Yuliang Wu, Wei Zhai, Yuxin Cui, Tiesong Zhao, Yang Cao, Zheng-Jun Zha
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 neighborhood-greedy optimization (well-converged subregions guide poorly-converged neighbors) (ablation) | AIS framework · Event-VDM benchmark, 100+ pixel displacements | survival rate: 39.0% -> 65.7% (Table 3); text states 49.0% -> 87.1% | Improved |
| add neighborhood-greedy optimization with convergence assessment (ablation) | AIS framework · Event-VDM benchmark, 100+ pixel displacements | mean convergence time: 26.5s -> 7.2s | Improved |
| use affine-invariant displacement interpolation instead of nearest neighbor, mean, IDW or Gaussian weighting (ablation) | AIS framework · Event-VDM benchmark | EPE and survival rate | Improved |
| fuse events with 5 fps frames instead of OpenCorr on 5 fps frames only | Ours · Event-VDM benchmark, 100+ pixel displacements | survival rate: 25.0% -> 65.7% | Improved |
| add events to frames instead of TimeLens interpolated frames for image-based baselines | Ours · Event-VDM benchmark, 100+ pixel displacements | survival rate: 41.3% -> 65.7% (vs OpenCorr + TimeLens) | Improved |
| add TimeLens frame interpolation to OpenCorr | OpenCorr · Event-VDM benchmark, 100+ pixel displacements | survival rate: 25.0% -> 41.3% | Improved |
| add TimeLens frame interpolation to CoTrackerV3 | CoTrackerV3 · Event-VDM benchmark, 100+ pixel displacements | survival rate: 45.2% -> 29.1% | Hurt |
| add TimeLens frame interpolation to StrainNet | StrainNet · Event-VDM benchmark, 100+ pixel displacements | survival rate: 18.8% -> 35.4% | Improved |
| increase frame rate paired with events from 1 to 5 fps (ablation) | Ours · Event-VDM benchmark, 100+ pixel displacements | survival rate: 32.6% -> 65.7% | Improved |
| increase frame rate paired with events from 5 to 10 fps (ablation) | Ours · Event-VDM benchmark, 100+ pixel displacements | survival rate: 65.7% -> 67.5% | Improved |
| increase frame rate paired with events from 10 to 20 fps (ablation) | Ours · Event-VDM benchmark, 100+ pixel displacements | survival rate: 67.5% -> 71.2% | Improved |
| use 5 fps frames with events instead of 100 fps high-speed frames | Ours vs OpenCorr · Event-VDM benchmark, 100+ pixel displacements | survival rate: 64.3% (100 fps OpenCorr) vs 65.7% | No clear effect |
Closest papers by meaning
- Event-Based Non-rigid Reconstruction of Low-Rank Parametrized Deformations from Contours (IJCV 2024 · similarity 0.88)
- EventVGGT: Exploring Cross-Modal Distillation for Consistent Event-based Depth Estimation (ECCV 2026 · similarity 0.87)
- EMoTive: Event-guided Trajectory Modeling for 3D Motion Estimation (ICCV 2025 · similarity 0.87)
- Event-based Structure-from-Orbit (CVPR 2024 · similarity 0.85)
- TimeTracker: Event-based Continuous Point Tracking for Video Frame Interpolation with Non-linear Motion (CVPR 2025 · similarity 0.85)
- Event-Driven Dynamic Scene Depth Completion (NeurIPS 2025 · similarity 0.85)
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.
