Research map / Label-efficient and robust learning / Semi-supervised learning
Spatial-SAM: Spatially Consistent 3D Electron Microscopy Segmentation with SDF Memory and Semi-Supervised Learning
Yikai Huang, Renmin Han, Yuxuan Wang, Youcheng Cai, Ligang Liu
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 |
|---|---|---|---|
| replace SAM2 logit memory with SDF memory (unidirectional) (ablation) | Spatial-SAM · MitoEM-R (1/64 labels) | Dice: 92.62 -> 93.62 | Improved |
| make SDF memory bidirectional instead of unidirectional (ablation) | Spatial-SAM · MitoEM-R (1/64 labels) | Dice: 93.62 -> 94.45 | Improved |
| exclude the current slice's own SDF memory from the neighborhood (ablation) | Spatial-SAM · MitoEM-R (1/64 labels) | Dice: 94.11 -> 94.45 | Improved |
| increase labeled slices from 1/64 to 1/4 (ablation) | Spatial-SAM · MitoEM-R | Dice: 94.45 -> 95.35 | Improved |
| use Spatial-SAM with 1/64 labels instead of CPS U-Net | semi-supervised 3D EM segmentation · four-dataset average | Dice: +8.07% Dice, +11.49% mIoU | Improved |
| Spatial-SAM with 1/64 labels versus fully supervised SAM4EM | Spatial-SAM · four-dataset average | Dice: -0.09% Dice, -0.10% mIoU | No clear effect |
| more dual-track iterations (SDF and mask training) beyond three (ablation) | Spatial-SAM · validation mIoU | mIoU | No clear effect |
| SAM2-initialized pseudo-labels with dual-track refinement | Spatial-SAM · OOMLN (nuclei) | Dice: 92.91 -> 98.14 (vs 3D U-Net full labels) | Improved |
Closest papers by meaning
- $\epsilon$-Seg: Sparsely Supervised Semantic Segmentation of Microscopy Data (NeurIPS 2025 · similarity 0.86)
- Cross-Dimension Affinity Distillation for 3D EM Neuron Segmentation (CVPR 2024 · similarity 0.85)
- Electron Microscopy Images as Set of Fragments for Mitochondrial Segmentation (AAAI 2024 · similarity 0.85)
- A Gaussian Filter-Based 3D Registration Method for Series Section Electron Microscopy (AAAI 2025 · similarity 0.84)
- Alleviate and Mining: Rethinking Unsupervised Domain Adaptation for Mitochondria Segmentation from Pseudo-Label Perspective (AAAI 2025 · similarity 0.84)
- Evidential Uncertainty-Guided Mitochondria Segmentation for 3D EM Images (AAAI 2024 · similarity 0.84)
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.
