Research map / Image generation / Image restoration
DoPAMINE: Double-Sided Masked CNN for Pixel Adaptive Multiplicative Noise Despeckling
Sunghwan Joo, Sungmin Cha, Taesup Moon
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 DoPAMINE_S-AFT instead of SAR-DRN_S | UCML despeckling · UCML, L=1.0 | PSNR (dB): 22.86 -> 23.308 | Improved |
| Use DoPAMINE_S-AFT instead of SAR-DRN_S | UCML despeckling · UCML, L=2.0 | PSNR (dB): 24.189 -> 24.701 | Improved |
| Use DoPAMINE_S-AFT instead of SAR-DRN_S | UCML despeckling · UCML, L=4.0 | PSNR (dB): 25.664 -> 26.123 | Improved |
| Use DoPAMINE_S-AFT instead of SAR-DRN_S | UCML despeckling · UCML, L=8.0 | PSNR (dB): 27.191 -> 27.623 | Improved |
| Add adaptive fine-tuning to DoPAMINE_S | DoPAMINE_S on UCML · UCML, L=1.0–8.0 | PSNR (dB): 0.3~0.4dB | Improved |
| Add data augmentation to fine-tuning (ablation) | DoPAMINE on UCML · UCML, L=1.0 | PSNR: 0.3dB | Improved |
| Use a blind supervised model followed by AFT instead of the blind model alone | DoPAMINE_B on UCML · UCML, L=1.0–8.0 | PSNR | Improved |
| Add a final ResNet block (ablation) | DoPAMINE_S on UCML · UCML, L=1.0 | PSNR | Improved |
| Use the slope parameter a_i in the pixelwise affine mapping (ablation) | DoPAMINE on UCML · UCML, L=1.0 | PSNR | Improved |
| Use scale-add layers instead of ordinary addition with batch normalization (ablation) | DoPAMINE_S · UCML, L=1.0 | PSNR | Improved |
| Apply AFT to DoPAMINE_B on Flevoland | DoPAMINE_B · Flevoland SAR | Avg ENL: 74.1 -> 418.4 | Improved |
| Use DoPAMINE_B-AFT instead of SAR-DRN_S | Flevoland SAR despeckling · Flevoland SAR | Avg ENL: 381.7 -> 418.4 | Improved |
Closest papers by meaning
- GAN2GAN: Generative Noise Learning for Blind Denoising with Single Noisy Images (ICLR 2021 · similarity 0.86)
- Dual Adversarial Network: Toward Real-world Noise Removal and Noise Generation (ECCV 2020 · similarity 0.85)
- Universal Denoising Networks : A Novel CNN Architecture for Image Denoising (CVPR 2018 · similarity 0.85)
- Unpaired Learning of Deep Image Denoising (ECCV 2020 · similarity 0.85)
- Dual-Perspective Activation: Efficient Channel Denoising via Joint Forward-Backward Criterion for Artificial Neural Networks (NeurIPS 2024 · similarity 0.84)
- Zero-Shot Noise2Mean: Gap Minimization for Efficient Denoising from a Single Noisy Image (AAAI 2025 · similarity 0.83)
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.
