Research map / 3D vision / Face and shape
FPicker: Topology-Guided Evolution for Filament Tracing in Low-SNR Microscopy
Tingyin Zhao, Mingtao Huang, Yuan Shen
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 box prior with center-endpoint representation (ablation) | FPicker (closed curve) · Cryo-Sim Extreme SNR | mSAP (%): 2.5 -> 12.5 | Improved |
| replace closed with open-boundary graph convolution (ablation) | FPicker · Cryo-Sim Medium SNR | P-Ang (deg): 45.2 -> 19.4 | Improved |
| open-curve evolution on real data (ablation) | FPicker · Custom-EMPIAR (fine-tuned) | P-Ang (deg): 54.8 -> 42.3 | Improved |
| open-curve evolution F1 effect (ablation) | FPicker · Cryo-Sim Medium SNR | F1 (%): 73.2 -> 75.8 | Improved |
| add auxiliary geometric field loss (ablation) | FPicker · Cryo-Sim Extreme SNR | mSAP (%): 35.6 -> 45.2 | Improved |
| add bridge curriculum strategy (ablation) | FPicker · Cryo-Sim Extreme SNR | mSAP (%): 45.2 -> 76.5 | Improved |
| add bridge curriculum strategy (ablation) | FPicker · Custom-EMPIAR (fine-tuned) | mSAP (%): 71.4 -> 82.9 | Improved |
| increase deformation iterations from 1 to 3 (ablation) | FPicker · Cryo-Sim Medium SNR | mSAP (%): 78.5 -> 91.1 | Improved |
| DLA-34 backbone vs ResNet-50 backbone (ablation) | FPicker · Cryo-Sim Extreme SNR | mSAP (%): 72.1 -> 76.5 | Improved |
| fine-tune on real EMPIAR images vs zero-shot | FPicker · Custom-EMPIAR | mSAP (%): 28.6 -> 82.9 | Improved |
| more real images for fine-tuning (5 to 200) (ablation) | FPicker · Custom-EMPIAR | mSAP (%): 41.3 -> 80.4 | Improved |
| box-based detectors under extreme noise vs high SNR | YOLOv8 · Cryo-Sim | mSAP (%): 78.5 -> 28.5 | Hurt |
Closest papers by meaning
- TopoFuse: Topology-Aware Tri-Planar Fusion for 3D Cryo-Electron Tomography Segmentation (ECCV 2026 · similarity 0.89)
- SkelEM: Explicit Decoupling of Topology and Details for Self-supervised Axial Super-Resolution in Volume Microscopy (ECCV 2026 · similarity 0.85)
- SILSA: Sliding-Window Slice Latents for Topology-Preserving High-Resolution 3D Generation (NeurIPS 2026 · similarity 0.84)
- CryoSPIN: Improving Ab-Initio Cryo-EM Reconstruction with Semi-Amortized Pose Inference (NeurIPS 2024 · similarity 0.83)
- ProtPainter: Draw or Drag Protein via Topology-guided Diffusion (ICLR 2025 · similarity 0.83)
- At FullTilt: Real-Time Open-Set 3D Macromolecule Detection Directly from Tilted 2D Projections (NeurIPS 2026 · similarity 0.83)
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.
