Research map / 3D vision / Depth and stereo
Spike Camera Autofocus via Frequency-Domain Spectral-Centroid Migration
Xijie Xiang, Lin Zhu, 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 |
|---|---|---|---|
| use CEN instead of best baseline SD | autofocus · SAD real spike sequences | mean absolute error (blocks): 6.67 -> 2.80 | Improved |
| use CEN on low-light synthetic sweeps instead of HF | autofocus · CVIA synthetic (low-light) | mean error (cm): 1.90 -> 1.30 | Improved |
| use CEN on moderate-light synthetic sweeps | autofocus · CVIA synthetic (moderate-light) | mean error (cm): 0.24 -> 0.00 | Improved |
| operate directly on spikes instead of reconstructing images (TFI/count) before GRAD/LAP/RDF/DRDF | CEN · SAD real spike sequences | mean absolute error (blocks): 28.20 -> 2.80 | Improved |
| use CEN instead of SD in static scenes | autofocus · SAD real spike sequences (static) | AbsErr: 3.00 vs 3.14 | No clear effect |
| use CEN instead of SD in dynamic scenes | autofocus · SAD real spike sequences (dynamic) | AbsErr: 9.50 -> 2.50 | Improved |
| use CEN instead of HF on USAF | autofocus · SAD real spike sequences USAF scene | AbsErr: 3.33 -> 3.44 | Hurt |
| add FFT shift to DC removal (ablation) | CEN · SAD real spike sequences | mean absolute error (blocks): 35.47 -> 2.80 | Improved |
| add DC removal to FFT shift (ablation) | CEN · SAD real spike sequences | mean absolute error (blocks): 6.67 -> 2.80 | Improved |
| add structure-consistency curve selection with plateau, edge, curvature terms to prominence+width (ablation) | CEN · SAD real spike sequences | mean absolute error (blocks): 33.40 -> 2.80 | Improved |
| add prominence term (ablation) | CEN · SAD real spike sequences | mean absolute error (blocks): 35.27 -> 2.80 | Improved |
| add edge position term (ablation) | CEN · SAD real spike sequences | mean absolute error (blocks): 33.07 -> 2.80 | Improved |
Closest papers by meaning
- Intensity-Robust Autofocus for Spike Camera (CVPR 2024 · similarity 0.87)
- SpikeNeRF: Learning Neural Radiance Fields from Continuous Spike Stream (CVPR 2024 · similarity 0.82)
- Finding Visual Saliency in Continuous Spike Stream (AAAI 2024 · similarity 0.82)
- Boosting Spike Camera Image Reconstruction from a Perspective of Dealing with Spike Fluctuations (CVPR 2024 · similarity 0.82)
- One-Step Event-Driven High-Speed Autofocus (CVPR 2025 · similarity 0.81)
- EvFocus: Learning to Reconstruct Sharp Images from Out-of-Focus Event Streams (ICML 2025 · similarity 0.81)
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.
