Research map / Video / Video segmentation and tracking
Unsupervised Deep Video Denoising with Untrained Network
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 VER2R instead of R2R for real-noise training | The proposed video denoiser · Real raw video, mean | PSNR (dB): 44.70 -> 45.64 | Helped |
| Add motion correction (ablation) | The proposed video denoiser · DAVIS, σ=20 and σ=40 | PSNR (dB): 36.17 -> 36.49; 33.18 -> 33.29 | Helped |
| Add temporal attention (ablation) | The proposed video denoiser · DAVIS, σ=20 and σ=40 | PSNR (dB): 35.36 -> 36.49; 32.70 -> 33.29 | Helped |
| Use the full spatial-temporal denoiser rather than only the spatial denoiser (ablation) | The proposed video denoiser · DAVIS, σ=20 and σ=40 | PSNR (dB): 34.77 -> 36.49; 31.91 -> 33.29 | Helped |
| Use 5 adjacent frames rather than 3 (ablation) | The proposed video denoiser · DAVIS, σ=20 and σ=40 | PSNR (dB): 36.09 -> 36.49; 33.04 -> 33.29 | Helped |
| Replace temporal-only fusion with 3D spatial-temporal attention kernels (ablation) | The proposed video denoiser · DAVIS, σ=20 and σ=40 | PSNR (dB): 36.49 -> 36.37; 33.29 -> 33.30 | Mixed |
Closest papers by meaning
- Classic Video Denoising in a Machine Learning World: Robust, Fast, and Controllable (CVPR 2025 · similarity 0.88)
- An Unsupervised Deep Learning Approach for Real-World Image Denoising (ICLR 2021 · similarity 0.87)
- Model-Blind Video Denoising via Frame-To-Frame Training (CVPR 2019 · similarity 0.87)
- FastDVDnet: Towards Real-Time Deep Video Denoising Without Flow Estimation (CVPR 2020 · similarity 0.86)
- IDR: Self-Supervised Image Denoising via Iterative Data Refinement (CVPR 2022 · similarity 0.86)
- High-Quality Self-Supervised Deep Image Denoising (NeurIPS 2019 · 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.