Research map / Image generation / Image restoration
Modeling Mask Uncertainty in Hyperspectral Image Reconstruction
Jiamian Wang, Yulun Zhang, Xin Yuan, Ziyi Meng, Zhiqiang Tao
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 |
|---|---|---|---|
| add the GST network to SRN trained with mask ensemble (ablation) | SRN · CASSI simulation, 10 scenes (many-to-many miscalibration) | PSNR (dB): 32.24 -> 33.02 | Improved |
| add bilevel optimization (ablation) | GST · CASSI simulation, 10 scenes (many-to-many miscalibration) | PSNR (dB): 32.43 -> 33.02 | Improved |
| add GCN module in GST instead of convolution (ablation) | GST · CASSI simulation, 10 scenes (many-to-many miscalibration) | PSNR (dB): 32.82 -> 33.02 | Improved |
| add bilevel optimization (ablation) | GST · CASSI simulation, 10 scenes (one-to-one, same mask) | PSNR (dB): 32.73 -> 33.22 | Improved |
| add GST network to SRN (ablation) | SRN · CASSI simulation, 10 scenes (one-to-one, same mask) | PSNR (dB): 33.17 -> 33.22 | No clear effect |
| add GCN module instead of convolution (ablation) | GST · CASSI simulation, 10 scenes (one-to-one, same mask) | PSNR (dB): 33.23 -> 33.22 | No clear effect |
| add GST network to SRN (ablation) | SRN · CASSI simulation, 10 scenes (one-to-many miscalibration) | PSNR (dB): 30.17 -> 30.60 | Improved |
| add bilevel optimization (ablation) | GST · CASSI simulation, 10 scenes (one-to-many miscalibration) | PSNR (dB): 30.30 -> 30.60 | Improved |
| add GCN module instead of convolution (ablation) | GST · CASSI simulation, 10 scenes (one-to-many miscalibration) | PSNR (dB): 30.13 -> 30.60 | Improved |
| add GST to Spectral ViT backbone (ablation) | Spectral ViT · CASSI simulation, 10 scenes (many-to-many), backbone comparison | PSNR (dB): 31.62 -> 32.15 | Improved |
| add GST to SwinIR backbone (ablation) | SwinIR · CASSI simulation, 10 scenes (many-to-many), backbone comparison | PSNR (dB): 33.49 -> 34.15 | Improved |
| train on mask ensemble instead of single mask | TSA-Net · CASSI simulation, 10 scenes | PSNR (dB): 27.47 -> 21.42 | Hurt |
Closest papers by meaning
- Mask-Guided Spectral-Wise Transformer for Efficient Hyperspectral Image Reconstruction (CVPR 2022 · similarity 0.92)
- Uncertainty-Driven Spectral Compressive Imaging with Spatial-Frequency Transformer (ECCV 2024 · similarity 0.90)
- S2-Transformer for Mask-Aware Hyperspectral Image Reconstruction (TPAMI 2025 · similarity 0.89)
- Deep Gaussian Scale Mixture Prior for Spectral Compressive Imaging (CVPR 2021 · similarity 0.89)
- Dual-Window Multiscale Transformer for Hyperspectral Snapshot Compressive Imaging (AAAI 2024 · similarity 0.88)
- Improving Spectral Snapshot Reconstruction with Spectral-Spatial Rectification (CVPR 2024 · similarity 0.87)
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.
