Research map / Optimization / Optimizers
Deep Gradient Projection Networks for Pan-sharpening
Check 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 GPPNN instead of PanNet | Pan-sharpening · Landsat8 test | PSNR: 38.0910 -> 38.9939 | Helped |
| use GPPNN instead of MHNet | Pan-sharpening · QuickBird test | PSNR: 31.1557 -> 31.4973 | Helped |
| use GPPNN instead of DRPNN | Pan-sharpening · GaoFen2 test | PSNR: 35.1182 -> 35.9680 | Helped |
| add proximal modules that impose learned deep priors (ablation) | GPPNN · Landsat8 | PSNR: 37.0404 -> 38.9939 | Helped |
| use distinct weights rather than sharing weights across layers (ablation) | GPPNN · Landsat8 | PSNR: 38.1650 -> 38.9939 | Helped |
| use separate LRMS-aware and PAN-aware subproblems rather than one combined-problem network (ablation) | GPPNN · Landsat8 | PSNR: 38.3213 -> 38.9939 | Helped |
| use untied convolution kernels rather than kernels constrained to be transposes (ablation) | GPPNN · Landsat8 | PSNR: 38.5487 -> 38.9939 | Helped |
| increase depth from 2 to 14 layers (ablation) | GPPNN with C=64 · Landsat8 validation | PSNR: 39.0648 -> 40.0509 | Mixed |
| increase width from 8 to 128 filters (ablation) | GPPNN with K=8 · Landsat8 validation | PSNR: 36.6455 -> 39.0841 | Mixed |
Closest papers by meaning
- Pan-LUT: Efficient Pan-sharpening via Learnable Look-Up Tables (NeurIPS 2025 · similarity 0.88)
- Binarized Neural Network for Multi-spectral Image Fusion (CVPR 2025 · similarity 0.88)
- Unsupervised Pan-Sharpening via Mutually Guided Detail Restoration (AAAI 2024 · similarity 0.86)
- SSUN-Net: Spatial-Spectral Prior-Aware Unfolding Network for Pan-Sharpening (AAAI 2025 · similarity 0.85)
- Progressive High-Frequency Reconstruction for Pan-Sharpening with Implicit Neural Representation (AAAI 2024 · similarity 0.85)
- Frequency-Adaptive Pan-Sharpening with Mixture of Experts (AAAI 2024 · similarity 0.85)
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.