Research map / Image generation / Super-resolution
Pixel-Aware Deep Function-Mixture Network for Spectral Super-Resolution
Lei Zhang, Zhiqiang Lang, Peng Wang, Wei Wei, Shengcai Liao, Ling Shao, Yanning Zhang
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 pixel-wise mixture weights rather than equal weights for basis functions (ablation) | Pixel-aware function-mixture network · NTIRE2018 | PSNR: 48.44 -> 49.29 | Improved |
| Add intermediate feature fusion with skip connections and a fusion FM block (ablation) | Pixel-aware function-mixture network · NTIRE2018 | PSNR: 48.97 -> 49.29 | Improved |
| Increase basis-function count from 1 to 3 (ablation) | Pixel-aware function-mixture network · NTIRE2018 | PSNR: 45.82 -> 49.29 | Improved |
| Increase basis-function count from 2 to 3 (ablation) | Pixel-aware function-mixture network · NTIRE2018 | PSNR: 48.76 -> 49.29 | Improved |
| Increase basis-function count from 3 to 5 (ablation) | Pixel-aware function-mixture network · NTIRE2018 | PSNR: 49.29 -> 49.87 | Improved |
| Increase FM-block count from 2 to 3 (ablation) | Pixel-aware function-mixture network · NTIRE2018 | PSNR: 48.95 -> 49.29 | Improved |
| Increase FM-block count from 3 to 4 (ablation) | Pixel-aware function-mixture network · NTIRE2018 | PSNR: 49.29 -> 49.42 | Improved |
| Increase FM-block count from 3 to 6 (ablation) | Pixel-aware function-mixture network · NTIRE2018 | PSNR: 49.29 -> 49.59 | Improved |
| Replace the convolutional-block network with pixel-aware function-mixture blocks | DCNN baseline · NTIRE2018 | PSNR: 47.40 -> 49.29 | Improved |
| Replace the multi-column architecture with the proposed stacked function-mixture architecture | MCNet baseline · NTIRE2018 | PSNR: 48.43 -> 49.29 | Improved |
| Use the proposed pixel-aware function-mixture network rather than HSCNN+ | HSCNN+ · CAVE | RMSE: 4.97 -> 4.54 | Improved |
| Use the proposed pixel-aware function-mixture network rather than HSCNN+ | HSCNN+ · CAVE | PSNR: +0.67db | Improved |
Closest papers by meaning
- Hyperspectral Image Super-Resolution With Optimized RGB Guidance (CVPR 2019 · similarity 0.90)
- Unsupervised Sparse Dirichlet-Net for Hyperspectral Image Super-Resolution (CVPR 2018 · similarity 0.90)
- Joint Camera Spectral Sensitivity Selection and Hyperspectral Image Recovery (ECCV 2018 · similarity 0.89)
- EigenSR: Eigenimage-Bridged Pre-Trained RGB Learners for Single Hyperspectral Image Super-Resolution (AAAI 2025 · similarity 0.89)
- Cross-Attention in Coupled Unmixing Nets for Unsupervised Hyperspectral Super-Resolution (ECCV 2020 · similarity 0.89)
- HDNet: High-Resolution Dual-Domain Learning for Spectral Compressive Imaging (CVPR 2022 · similarity 0.89)
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.
