Research map / Image generation / Image restoration
DDS2M: Self-Supervised Denoising Diffusion Spatio-Spectral Model for Hyperspectral Image Restoration
Yuchun Miao, Lefei Zhang, Liangpei Zhang, Dacheng 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 diffusion process around the untrained spatial and spectral networks (ablation) | DDS2M · HSI denoising, Fruits (sigma 0.3) | PSNR: 31.983 -> 33.045 | Improved |
| add the diffusion process around the untrained spatial and spectral networks (ablation) | DDS2M · Noisy HSI completion, Fruits (sampling rate 0.1) | PSNR: 31.562 -> 33.217 | Improved |
| add the diffusion process around the untrained spatial and spectral networks (ablation) | DDS2M · Noisy HSI super-resolution, Fruits (x2) | PSNR: 31.172 -> 34.066 | Improved |
| use separate untrained spatial (U-Net) and spectral (FCN) networks instead of one untrained U-Net (ablation) | DDS2M · HSI denoising, Fruits (sigma 0.3) | PSNR: 29.682 -> 33.045 | Improved |
| use separate untrained spatial (U-Net) and spectral (FCN) networks instead of one untrained U-Net (ablation) | DDS2M · Noisy HSI completion, Fruits (sampling rate 0.1) | PSNR: 28.565 -> 33.217 | Improved |
| use separate untrained spatial (U-Net) and spectral (FCN) networks instead of one untrained U-Net (ablation) | DDS2M · Noisy HSI super-resolution, Fruits (x2) | PSNR: 32.588 -> 34.066 | Improved |
| replace pretrained diffusion denoiser (DDRM) with self-supervised untrained VS2M | DDRM · Noisy HSI completion, CAVE (sampling rate 0.1) | PSNR: 24.474 -> 32.507 | Improved |
| replace pretrained diffusion denoiser (DDRM) with self-supervised untrained VS2M | DDRM · Noisy HSI super-resolution, CAVE (x8) | PSNR: 18.883 -> 23.621 | Improved |
| use self-supervised DDS2M instead of supervised SST trained on ICVL (distribution shift) | HSI denoising, CAVE · HSI denoising, CAVE (sigma 0.3) | PSNR: 20.866 -> 32.951 | Improved |
| train supervised SR models on a different noise level or dataset than test | SFCSR/RFSR/SSPSR · Noisy HSI super-resolution, CAVE (x2, sigma 0.1) | PSNR: 30.350 -> 28.015 (SFCSR 0.1 -> 0.1*) | Hurt |
| increase sampling rate from 0.1 to 0.3 for DDNM | DDNM · Noisy HSI completion, CAVE | PSNR: 16.718 -> 32.222 | Improved |
Later papers that compared with this method
3 later accepted papers report DDS2M as a baseline in their results tables; 2 of them report a better number on most of the results they share with it.
- Deep Rank-One Tensor Functional Factorization for Multi-Dimensional Data Recovery (AAAI 2025 · better on 1 of 1 shared result)
- HIR-Diff: Unsupervised Hyperspectral Image Restoration Via Improved Diffusion Models (CVPR 2024 · better on 2 of 3 shared results)
- Unmixing Diffusion for Self-Supervised Hyperspectral Image Denoising (CVPR 2024 · better on 1 of 2 shared results)
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- HIR-Diff: Unsupervised Hyperspectral Image Restoration Via Improved Diffusion Models (CVPR 2024 · similarity 0.91)
- Latent Diffusion Enhanced Rectangle Transformer for Hyperspectral Image Restoration (TPAMI 2024 · similarity 0.91)
- Unmixing Diffusion for Self-Supervised Hyperspectral Image Denoising (CVPR 2024 · similarity 0.91)
- Residual Denoising Diffusion Models (CVPR 2024 · similarity 0.88)
- Enhancing Hyperspectral Images via Diffusion Model and Group-Autoencoder Super-resolution Network (AAAI 2024 · similarity 0.88)
- S2CycleDiff: Spatial-Spectral-Bilateral Cycle-Diffusion Framework for Hyperspectral Image Super-resolution (AAAI 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.
