Research map / Label-efficient and robust learning / Self-supervised and contrastive
LG-BPN: Local and Global Blind-Patch Network for Self-Supervised Real-World Denoising
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 |
|---|---|---|---|
| Add a 9×9 DSPMC local branch (ablation) | LG-BPN · SIDD validation | PSNR: 35.82 -> 37.32 | Helped |
| Add a 21×21 DSPMC global branch (ablation) | LG-BPN · SIDD validation | PSNR: 36.21 -> 37.32 | Helped |
| Use dilated convolution in the local branch rather than DTB (ablation) | LG-BPN · SIDD validation | PSNR: 36.90 -> 37.32 | Helped |
| Use DTB in the global branch rather than dilated convolution (ablation) | LG-BPN · SIDD validation | PSNR: 36.82 -> 37.32 | Helped |
| Add the kernel shift strategy at test time (ablation) | LG-BPN · SIDD validation | PSNR: 35.64 -> 37.32 | Helped |
| Use dilation 1 in the 9×9 branch and dilation 2 in the 21×21 branch (ablation) | LG-BPN DSPMC branches · SIDD validation | PSNR: 37.23 -> 37.32 | Helped |
| Increase 9×9 DSPMC dilation from 1 to 2 (ablation) | LG-BPN · SIDD validation | PSNR: 37.32 -> 36.84 | Hurt |
| Use dilation 3 in the 21×21 DSPMC branch (ablation) | LG-BPN · SIDD validation | PSNR: 37.32 -> 37.23 | Hurt |
Closest papers by meaning
- TM-BSN: Triangular-Masked Blind-Spot Network for Real-World Self-Supervised Image Denoising (CVPR 2026 · similarity 0.93)
- Exploring Efficient Asymmetric Blind-Spots for Self-Supervised Denoising in Real-World Scenarios (CVPR 2024 · similarity 0.91)
- APR-RD: Complemental Two Steps for Self-Supervised Real Image Denoising (AAAI 2025 · similarity 0.91)
- Rethinking Transformer-Based Blind-Spot Network for Self-Supervised Image Denoising (AAAI 2025 · similarity 0.90)
- Pseudo-Siamese Blind-spot Transformers for Self-Supervised Real-World Denoising (NeurIPS 2024 · similarity 0.89)
- Zero-Shot Blind-spot Image Denoising via Implicit Neural Sampling (CVPR 2025 · 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.