Research map / Graph learning / GNN architectures
MAVEN: A Mesh-Aware Volumetric Encoding Network for Simulating 3D Flexible Deformation
Zhe Feng, Shilong Tao, Haonan Sun, Shaohan Chen, Zhanxing Zhu, Yunhuai 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 |
|---|---|---|---|
| add geometry-aware aggregation coefficients (vs degree averaging) (ablation) | MAVEN · Cavity Grasping, full rollout | Position RMSE: 17.45 -> 15.41 | Improved |
| add geometry-aware aggregation coefficients (vs degree averaging) (ablation) | MAVEN · Metal Bending, full rollout | Stress RMSE: 6683.94 -> 4776.72 | Improved |
| add explicit cell/facet geometric input features (vs zero padding) (ablation) | MAVEN · Metal Bending, full rollout | Position RMSE: 1652.31 -> 810.42 | Improved |
| add explicit cell/facet geometric input features (vs zero padding) (ablation) | MAVEN · Cavity Grasping, full rollout | Position RMSE: 15.933 -> 15.41 | Improved |
| add explicit cell and facet nodes (vs averaging features onto vertices) (ablation) | MAVEN · Metal Bending, full rollout | Stress RMSE: 10375.86 -> 4776.72 | Improved |
| add explicit cell and facet nodes (vs averaging features onto vertices) (ablation) | MAVEN · Metal Bending, full rollout | Position RMSE: 1680.20 -> 810.42 | Improved |
| add cell nodes on top of facet-only FIGNet | MAVEN vs FIGNet · Metal Bending, full rollout | Position RMSE: 1030.57 -> 810.42 | Improved |
| use MAVEN instead of best node-based baseline as mesh gets coarser | MAVEN · DP, CG, MBD | average improvement: 3.41%, 13.07%, and 18.13% | Improved |
| use MAVEN instead of MGN | MAVEN · Deforming Plate, full rollout | Position RMSE: 23.65 -> 23.41 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- HOOD — HOOD: Hierarchical Graphs for Generalized Modelling of Clothing Dynamics (CVPR 2023) · better on 4 of 4 shared results
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- Diffusion-Based Hierarchical Graph Neural Networks for Simulating Nonlinear Solid Mechanics (NeurIPS 2025 · similarity 0.88)
- Neural Shape Deformation Priors (NeurIPS 2022 · similarity 0.86)
- MMGP: a Mesh Morphing Gaussian Process-based machine learning method for regression of physical problems under nonparametrized geometrical variability (NeurIPS 2023 · similarity 0.86)
- Pixel2Mesh: 3D Mesh Model Generation via Image Guided Deformation (TPAMI 2020 · similarity 0.86)
- Thickness-aware E(3)-Equivariant 3D Mesh Neural Networks (ICML 2025 · similarity 0.85)
- Implicit-ARAP: Efficient Handle-Guided Neural Field Deformation via Local Patch Meshing (NeurIPS 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.
