Research map / Video / Video segmentation and tracking
Formulating Event-based Image Reconstruction as a Linear Inverse Problem with Deep Regularization using Optical Flow
Zelin Zhang, Anthony Joseph Yezzi, Guillermo Gallego
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 CNN denoiser regularizer instead of prior Laplacian solver | Laplacian reconstruction · brain sequence | SSIM: 0.7130 -> 0.7394 | Improved |
| use CNN denoiser regularizer instead of prior Laplacian solver | Laplacian reconstruction · brain sequence | MSE: 0.0084 -> 0.0102 | Hurt |
| use CNN denoiser regularizer instead of prior Laplacian solver | Laplacian reconstruction · chair sequence | SSIM: 0.8307 -> 0.8246 | Hurt |
| use CNN regularizer instead of TV (ablation) | Ours · N-Caltech accordion (unequalized) | MSE: 0.0365205 -> 0.03307699 | Improved |
| use CNN regularizer instead of Tikhonov (ablation) | Ours · N-Caltech accordion (unequalized) | SSIM: 0.5807396 -> 0.63323015 | Improved |
| evaluate without histogram equalization | Ours (CNN) · N-Caltech accordion | MSE: 0.04423198 -> 0.03307699 | Improved |
| increase regularizer weight lambda from 0.23 to 2.3 (ablation) | Ours (CNN) · qualitative | visual quality | Hurt |
| decrease regularizer weight lambda to 0.023 (ablation) | Ours (CNN) · qualitative | visual quality | Mixed |
| add uniform noise to optical flow beyond 2 pixels (ablation) | Ours (CNN, TV, Tikhonov) · boxes scene | SSIM and LPIPS | Hurt |
| use flow noise up to 1 pixel (ablation) | Ours · boxes scene | visual quality | No clear effect |
| use CNN regularizer instead of Tikhonov or TV (ablation) | Ours · runtime | seconds: 0.4401 -> 28.3904 | Hurt |
Closest papers by meaning
- Learning To Reconstruct High Speed and High Dynamic Range Videos From Events (CVPR 2021 · similarity 0.90)
- Events-To-Video: Bringing Modern Computer Vision to Event Cameras (CVPR 2019 · similarity 0.90)
- High Speed and High Dynamic Range Video with an Event Camera (TPAMI 2019 · similarity 0.90)
- Joint Framework for Single Image Reconstruction and Super-Resolution With an Event Camera (TPAMI 2021 · similarity 0.89)
- Simultaneous Optical Flow and Intensity Estimation From an Event Camera (CVPR 2016 · similarity 0.89)
- LED: A Large-scale Real-world Paired Dataset for Event Camera Denoising (CVPR 2024 · similarity 0.87)
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.
