Research map / Video / Video segmentation and tracking
Event Structural Valley: A Unified Theoretical and Practical Framework for Event Camera Autofocus
Xijie Xiang, Lin Zhu, Wei Zhang, Yonghong Tian
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 |
|---|---|---|---|
| locate focus at the event-rate valley between two peaks instead of the maximum event rate | ESVA vs ER+EGS · SYN average | Focus timestamp error (ms): 36.10 -> 6.62 | Improved |
| locate focus at the event-rate valley instead of the maximum | ESVA vs ER+EGS · EAD average | Focus distance error (um): 92.30 -> 65.38 | Improved |
| valley-based ESVA versus best prior method ELP | ESVA · DAVIS average | Focus timestamp error (ms): 2.04 -> 1.30 | Improved |
| valley-based ESVA versus best prior method ELP | ESVA · EVK4 average | Focus timestamp error (ms): 5.33 -> 4.22 | Improved |
| ESVA under strong shake | ESVA · SYN Huge Shake | Focus timestamp error (ms): ELP 9.43 vs ESVA 11.25 | Hurt |
| ESVA in dark shake scenes | ESVA · EAD Dark Shake | Focus distance error (um): OLE'23 53.57 vs ESVA 108.29 | Mixed |
| add structural (Gaussian) smoothing of the event-rate curve (ablation) | ESVA · DAVIS focusboard-bright-motion | Focus timestamp error (ms): > 43 ms without smoothing | Improved |
| add consistency filtering (ablation) | ESVA · EAD crossroad-dark-static | Focus distance error (um): 1,057 -> 54 | Improved |
| add dual-peak constraint (ablation) | ESVA · DAVIS ghost-bright-motion | Focus error | Improved |
| ESVA runtime versus ELP | ESVA · EVK4 runtime | Runtime (ms): 3699.67 -> 1.68 | Improved |
Closest papers by meaning
- EvFocus: Learning to Reconstruct Sharp Images from Out-of-Focus Event Streams (ICML 2025 · similarity 0.90)
- One-Step Event-Driven High-Speed Autofocus (CVPR 2025 · similarity 0.90)
- Autofocus for Event Cameras (CVPR 2022 · similarity 0.89)
- Focus Is All You Need: Loss Functions for Event-Based Vision (CVPR 2019 · similarity 0.88)
- Event Fields: Capturing Light Fields at High Speed, Resolution, and Dynamic Range (CVPR 2025 · similarity 0.87)
- Towards Robust Event-guided Low-Light Image Enhancement: A Large-Scale Real-World Event-Image Dataset and Novel Approach (CVPR 2024 · similarity 0.86)
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.
