Research map / Image generation / Image restoration
Self-Supervised Cryo-Electron Tomography Volumetric Image Restoration From Single Noisy Volume With Sparsity Constraint
Zhidong Yang, Fa Zhang, Renmin Han
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 up-sampling block before the encoder (ablation) | SC-Net · Simulated Synapse (sigma=15) | PSNR (dB): 22.47 -> 27.55 | Improved |
| add up-sampling block before the encoder (ablation) | SC-Net · Simulated Adhesion Belt (sigma=15) | PSNR (dB): 24.91 -> 25.68 | Improved |
| add expectancy constraint loss (ablation) | SC-Net · Simulated SARS-CoV (sigma=20) | SSIM: 0.780 -> 0.852 | Improved |
| add expectancy constraint loss (ablation) | SC-Net · Simulated Adhesion Belt (sigma=20) | PSNR (dB): 26.46 -> 23.41 | Hurt |
| add expectancy constraint loss (ablation) | SC-Net · Simulated SARS-CoV (sigma=20) | PSNR (dB): 27.58 -> 32.79 | Improved |
| add sparsity-guided smoothing loss (ablation) | SC-Net · Simulated data with insufficient training data | noise removal | Improved |
| replace Topaz 2D filter by BM3D for smoothed volume (ablation) | SC-Net · Simulated Synapse (sigma=10) | PSNR (dB): 30.03 -> 28.10 | Hurt |
| replace Topaz 2D filter by BM3D for smoothed volume (ablation) | SC-Net · Simulated Adhesion Belt (sigma=10) | PSNR (dB): 28.61 -> 33.30 | Improved |
| use sparsity-guided SC-Net instead of Noise2Void (both single-volume) | SC-Net · Simulated SARS-CoV (sigma=20) | PSNR (dB): 17.35 -> 32.79 | Improved |
| use SC-Net instead of Noise2Void on real data | SC-Net · Real Centriole | FSC e/o(0.5) resolution (Angstrom): 57.67 -> 38.21 | Improved |
| use SC-Net instead of Topaz (trained on hundreds of tomograms) | SC-Net · Real VEEV | FSC e/o(0.5) resolution (Angstrom): 6.8 -> 8.31 | Hurt |
| use SC-Net instead of Topaz | SC-Net · Simulated Synapse (sigma=20) | SSIM: 0.980 -> 0.860 | Hurt |
Closest papers by meaning
- DRACO: A Denoising-Reconstruction Autoencoder for Cryo-EM (NeurIPS 2024 · similarity 0.88)
- Vox-UDA: Voxel-wise Unsupervised Domain Adaptation for Cryo-Electron Subtomogram Segmentation with Denoised Pseudo-Labeling (AAAI 2025 · similarity 0.87)
- cryoSPHERE: Single-Particle HEterogeneous REconstruction from cryo EM (ICLR 2025 · similarity 0.86)
- At FullTilt: Real-Time Open-Set 3D Macromolecule Detection Directly from Tilted 2D Projections (NeurIPS 2026 · similarity 0.86)
- CryoSPIN: Improving Ab-Initio Cryo-EM Reconstruction with Semi-Amortized Pose Inference (NeurIPS 2024 · similarity 0.86)
- BOE-ViT: Boosting Orientation Estimation with Equivariance in Self-Supervised 3D Subtomogram Alignment (CVPR 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.
