Research map / Architectures and efficiency / Attention and transformers
End-to-End Low Cost Compressive Spectral Imaging with Spatial-Spectral Self-Attention
Ziyi Meng, Jiawei Ma, Xin Yuan
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 TSA-Net instead of lambda-net (ablation) | TSA-Net · KAIST simulation noise-free | PSNR: 29.25 -> 30.15 | Improved |
| use TSA-Net instead of U-net (ablation) | TSA-Net · KAIST simulation noise-free | PSNR: 26.80 -> 30.15 | Improved |
| use TSA-Net instead of HSSP | TSA-Net · KAIST simulation noise-free | PSNR: 28.93 -> 30.15 | Improved |
| use TSA-Net instead of DeSCI | TSA-Net · KAIST simulation noise-free | PSNR: 25.86 -> 30.15 | Improved |
| add shot noise to training data (ablation) | TSA-Net · KAIST 10-bit shot noise test | PSNR: 25.74 -> 28.08 | Improved |
| add shot noise to training data (ablation) | TSA-Net · KAIST 11-bit shot noise test | PSNR: 27.40 -> 28.35 | Improved |
| add shot noise to training data (ablation) | TSA-Net · KAIST 12-bit shot noise test | PSNR: 28.36 -> 28.55 | Improved |
| add shot noise to training data (ablation) | TSA-Net · KAIST noise-free test | PSNR: 30.15 -> 28.69 | Hurt |
| add shot noise to training data (ablation) | TSA-Net · KAIST Gaussian noise sigma=0.02 | PSNR: 22.65 -> 26.93 | Improved |
| add shot noise to training data (ablation) | TSA-Net · KAIST Gaussian noise sigma=0.05 | PSNR: 19.47 -> 23.50 | Improved |
| add shot noise to training data (ablation) | TSA-Net · KAIST Gaussian noise sigma=0.005 | PSNR: 28.33 -> 28.46 | No clear effect |
| add GAN training to the loss | TSA-Net · KAIST simulation | PSNR: 0.1dB | No clear effect |
Closest papers by meaning
- S2-Transformer for Mask-Aware Hyperspectral Image Reconstruction (TPAMI 2025 · similarity 0.92)
- Dual-Window Multiscale Transformer for Hyperspectral Snapshot Compressive Imaging (AAAI 2024 · similarity 0.91)
- Mask-Guided Spectral-Wise Transformer for Efficient Hyperspectral Image Reconstruction (CVPR 2022 · similarity 0.91)
- Spectral Compressive Imaging via Chromaticity-Intensity Decomposition (NeurIPS 2025 · similarity 0.91)
- Improving Spectral Snapshot Reconstruction with Spectral-Spatial Rectification (CVPR 2024 · similarity 0.89)
- SPECAT: SPatial-spEctral Cumulative-Attention Transformer for High-Resolution Hyperspectral Image Reconstruction (CVPR 2024 · 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.
