Research map / Image generation / Image restoration
Hyperspectral Image Fusion with Spectral-Band and Fusion-Scale Agnosticism
Yujie Liang, Zihan Cao, Liang-Jian Deng, Yang Yang, Malu Zhang
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 Matryoshka Kernel instead of zero-padding + mask in joint training (ablation) | SSA joint model · Pooled 7 datasets, overall | PSNR (dB): 37.64 -> 38.62 | Improved |
| Use Matryoshka Kernel instead of learned 1x1 spectral adapter (ablation) | SSA joint model · Pooled 7 datasets, overall | PSNR (dB): 37.92 -> 38.62 | Improved |
| Add MK with a single jointly trained model vs per-dataset w/o MK (ablation) | SSA · PaviaC x4 to x32 | PSNR (dB): 45.85 -> 45.92 at x4; 32.81 -> 32.79 at x32 | No clear effect |
| Joint training of one model across seven datasets instead of per-dataset models | SSA · CAVE and PaviaC x4 | PSNR (dB): 45.81 -> 45.96 on CAVE | Improved |
| Continuous INR decoding instead of x4 output resized by bicubic | SSA · CAVE x5.7 | PSNR (dB): 32.20 -> 43.49 | Improved |
| Use SSA instead of DCFormer at out-of-distribution scale | SSA · CAVE x8 | PSNR (dB): 41.46 vs 38.80 | Hurt |
| Fine-tune SSA for 500 iterations on unseen sensor | SSA · Houston (48 bands) x4 | PSNR (dB): 38.71 vs LRTN 38.65 | Improved |
| Fine-tune SSA for 500 iterations on unseen sensor | SSA · Loukia (176 bands) x4 | SAM: 3.95 vs DCFormer 3.88 | No clear effect |
Closest papers by meaning
- MHF-Net: An Interpretable Deep Network for Multispectral and Hyperspectral Image Fusion (TPAMI 2020 · similarity 0.89)
- Multispectral and Hyperspectral Image Fusion by MS/HS Fusion Net (CVPR 2019 · similarity 0.89)
- Deep Blind Hyperspectral Image Fusion (ICCV 2019 · similarity 0.88)
- A Selective Re-learning Mechanism for Hyperspectral Fusion Imaging (CVPR 2025 · similarity 0.87)
- Fourier-enhanced Implicit Neural Fusion Network for Multispectral and Hyperspectral Image Fusion (NeurIPS 2024 · similarity 0.87)
- OTIAS: OcTree Implicit Adaptive Sampling for Multispectral and Hyperspectral Image Fusion (AAAI 2025 · 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.
