Research map / 3D vision / Human pose estimation
A2-Net: Molecular Structure Estimation from Cryo-EM Density Volumes
Kui Xu, Zhe Wang, Jianping Shi, Hongsheng Li, Qiangfeng Cliff Zhang
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 conventional RoI pooling with APRoI (ablation) | A²-Net amino-acid detection · A² dataset | mAP: 0.610 -> 0.711 | Improved |
| use APRoI target size 16³ instead of 8³ (ablation) | A²-Net amino-acid detection · A² dataset | mAP | Improved |
| add sequence-guided neighbor loss (ablation) | A²-Net amino-acid detection and threading · A² dataset | mAP: 0.865 -> 0.891 | Improved |
| add sequence-guided neighbor loss (ablation) | A²-Net threading · A² dataset | Coverage: 0.72 -> 0.91 | Improved |
| use raw input cube rather than locNet feature volumes for RoI pooling (ablation) | 3D-ResNet detection · A² dataset | mAP: 0.416 -> 0.610 | Improved |
| add PBNet edge pruning (ablation) | MCTS threading · A² dataset | Coverage: 0.72 -> 0.91 | Improved |
| add PBNet edge pruning (ablation) | MCTS threading · A² dataset | RMSD: 2.9 -> 2.0 | Improved |
| use MCTS rather than distance-based DFS, with PBNet (ablation) | Threading · A² dataset | Coverage: 0.89 -> 0.91 | Improved |
| use MCTS rather than distance-based DFS, with PBNet (ablation) | Threading · A² dataset | RMSD: 2.6 -> 2.0 | Improved |
| pre-train 3D detection models on the A² dataset (ablation) | MV3D, Frustum-PointNet-v1, and Frustum-PointNet-v2 · KITTI car detection | AP: 65.53 -> 68.56; 58.97 -> 60.35; 59.14 -> 60.99; 83.26 -> 84.89; 69.28 -> 71.97; 62.56 -> 64.07; 83.76 -> 85.11; 70.92 -> | Improved |
| use A²-Net rather than Frustum-PointNet-v2 | Amino-acid detection · A² dataset | mAP: 0.425 -> 0.891 | Improved |
| use A²-Net rather than Frustum-PointNet-v2 | Amino-acid detection · A² dataset | Coverage: 0.48 -> 0.91 | Improved |
Closest papers by meaning
- cryoSPHERE: Single-Particle HEterogeneous REconstruction from cryo EM (ICLR 2025 · similarity 0.89)
- CryoSPIN: Improving Ab-Initio Cryo-EM Reconstruction with Semi-Amortized Pose Inference (NeurIPS 2024 · similarity 0.89)
- CryoFM: A Flow-based Foundation Model for Cryo-EM Densities (ICLR 2025 · similarity 0.89)
- CryoDomain: Sequence-free Protein Domain Identification from Low-resolution Cryo-EM Density Maps (AAAI 2025 · similarity 0.88)
- Reconstructing Heterogeneous Biomolecules via Hierarchical Gaussian Mixtures and Part Discovery (NeurIPS 2025 · similarity 0.88)
- Multiscale guidance of protein structure prediction with heterogeneous cryo-EM data (NeurIPS 2025 · similarity 0.88)
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.
