Research map / 3D vision / Depth and stereo
On the symmetries of the synchronization problem in Cryo-EM: Multi-Frequency Vector Diffusion Maps on the Projective Plane
Gabriele Cesa, Arash Behboodi, Taco Cohen, Max Welling
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 MFVDM O(2) denoising (L=5) to VDM synchronization (ablation) | VDM-based sync · Synthetic cryo-EM, N=5000, SNR=0.12 | pose correlation: 82.0 (0.69) -> 97.5 (0.22) | Improved |
| add MFVDM O(2) denoising (L=5) to VDM synchronization (ablation) | VDM-based sync · Synthetic cryo-EM, N=5000, SNR=0.05 | pose correlation: 60.1 (1.11) -> 63.4 (1.05) | Improved |
| add MFVDM O(2) denoising (L=5) to VDM synchronization (ablation) | VDM-based sync · Synthetic cryo-EM, N=3000, SNR=0.35 | pose correlation: 99.6 (0.1) -> 99.3 (0.13) | No clear effect |
| raise maximum frequency L from 5 to 10 in O(2) denoising (ablation) | VDM-based sync · Synthetic cryo-EM, N=5000, SNR=0.05 | pose correlation: 63.4 (1.05) -> 63.5 (1.05) | No clear effect |
| raise maximum frequency L from 5 to 10 in O(2) denoising (ablation) | CL synchronization · Synthetic cryo-EM, N=5000, SNR=0.05 | pose correlation: 75.9 (0.77) -> 76.6 (0.75) | Mixed |
| use O(2) instead of SO(2) denoising (L=5) (ablation) | CL synchronization · Synthetic cryo-EM, N=5000, SNR=0.05 | pose correlation: 71.4 (0.87) -> 75.9 (0.77) | Improved |
| use O(2) instead of SO(2) denoising (L=5) (ablation) | CL synchronization · RELION refinement, N=5000, SNR=0.05 | pose correlation: 59.4±11.5 -> 76.4±2.6 | Improved |
| use VDM synchronization instead of common lines (None denoising) | pose pipeline · RELION refinement, N=5000, SNR=0.05 | estimated resolution: 6.65±1.49 -> 3.51±0.06 | Improved |
| use VDM synchronization instead of common lines (None denoising) | pose pipeline · RELION refinement, N=5000, SNR=0.12 | pose correlation: 76.6 -> 99.9±0.0 | Improved |
| use CL instead of VDM synchronization | pose pipeline · Synthetic cryo-EM, N=5000, SNR=0.12 | pose correlation before refinement: 82.0 (0.69) -> 98.0 (0.20) | Improved |
| use O(2) relative poses instead of SO(2) in MFVDM (ablation) | MFVDM · synthetic vector dataset | correlation of cosine similarities | Improved |
| use higher maximum frequency L (multi-frequency) (ablation) | MFVDM O(2) · synthetic vector dataset | cosine similarity estimation | Improved |
Closest papers by meaning
- CryoBench: Diverse and challenging datasets for the heterogeneity problem in cryo-EM (NeurIPS 2024 · similarity 0.90)
- CryoSPIN: Improving Ab-Initio Cryo-EM Reconstruction with Semi-Amortized Pose Inference (NeurIPS 2024 · similarity 0.90)
- cryoSPHERE: Single-Particle HEterogeneous REconstruction from cryo EM (ICLR 2025 · similarity 0.90)
- Reconstructing Heterogeneous Biomolecules via Hierarchical Gaussian Mixtures and Part Discovery (NeurIPS 2025 · similarity 0.90)
- CryoFM: A Flow-based Foundation Model for Cryo-EM Densities (ICLR 2025 · similarity 0.89)
- Mixture of neural fields for heterogeneous reconstruction in cryo-EM (NeurIPS 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.
