Research map / Image generation / Super-resolution
Unsupervised Adaptation Learning for Hyperspectral Imagery Super-Resolution
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 |
|---|---|---|---|
| adapt the pre-trained fusion prior and estimate unknown degeneration jointly | UAL · CAVE, 40db noise, kernels k1-k4, s=8 | PSNR: 43.23 vs 40.26; 42.72 vs 39.53; 43.22 vs 39.62; 43.14 vs 39.53 | Helped |
| adapt the pre-trained fusion prior and estimate unknown degeneration jointly | UAL · CAVE, 30db and 35db noise, k1, s=8 | PSNR: 39.43 vs 37.16; 41.59 vs 38.17 | Helped |
| adapt the pre-trained fusion prior and estimate unknown degeneration jointly | UAL · CAVE, Harvard, and NTIRE, s=8 | PSNR: 43.23 vs 40.26; 42.38 vs 41.11; 46.19 vs 44.15 | Helped |
| train at scale s=32 and initialize the fusion module from the s=8 model | UAL · CAVE, Harvard, and NTIRE, s=32 | PSNR: 40.43 vs 37.65; 41.82 vs 40.63; 44.56 vs 43.26 | Helped |
| use UAL for real-data reconstruction | HypSen · HypSen real dataset | No-reference image quality score: 10.98 vs 11.19 | Helped |
| add the adaptation module (ablation) | UAL · CAVE, 40db, k1, s=8 | PSNR: 41.66 -> 43.23 | Helped |
| add the fusion module (ablation) | UAL · CAVE, 40db, k1, s=8 | PSNR: 36.17 -> 43.23 | Helped |
| pre-train the fusion module and keep its weights fixed during unsupervised learning (ablation) | UAL · CAVE, 40db, k1, s=8 | PSNR: 37.51 -> 43.23 | Helped |
| keep the pre-trained fusion module fixed instead of fine-tuning it (ablation) | UAL · CAVE, 40db, k1, s=8 | PSNR: 40.27 -> 43.23 | Helped |
| estimate degeneration instead of using pre-defined H-hat (ablation) | UAL · CAVE, 40db, k1, s=8 | PSNR: 38.58 -> 43.23 | Helped |
| use the multi-branch fusion inputs rather than only the concatenated input (ablation) | UAL · CAVE, 40db, k1, s=8 | PSNR: 37.42 -> 43.23 | Helped |
| remove the concatenation input branch (ablation) | UAL · CAVE, 40db, k1, s=8 | PSNR: 42.46 vs 43.23 | No clear effect |
Closest papers by meaning
- EigenSR: Eigenimage-Bridged Pre-Trained RGB Learners for Single Hyperspectral Image Super-Resolution (AAAI 2025 · similarity 0.89)
- Learning Multi-Modal Cross-Scale Deformable Transformer Network for Unregistered Hyperspectral Image Super-resolution (AAAI 2024 · similarity 0.88)
- HIR-Diff: Unsupervised Hyperspectral Image Restoration Via Improved Diffusion Models (CVPR 2024 · similarity 0.88)
- Enhancing Hyperspectral Images via Diffusion Model and Group-Autoencoder Super-resolution Network (AAAI 2024 · similarity 0.87)
- Unmixing Before Fusion: A Generalized Paradigm for Multi-Source-based Hyperspectral Image Synthesis (CVPR 2024 · similarity 0.87)
- S2CycleDiff: Spatial-Spectral-Bilateral Cycle-Diffusion Framework for Hyperspectral Image Super-resolution (AAAI 2024 · 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.