Research map / 3D vision / Face and shape
MoRE: 3D Visual Geometry Reconstruction Meets Mixture-of-Experts
Jingnan Gao, Zhe Wang, Xianze Fang, Xingyu Ren, Zhuo Chen, Shengqi Liu, Yuhao Cheng, Jiangjing Lyu, Xiaokang Yang, Yichao Yan
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 MoE layers (ablation) | MoRE (VGGT-based) · DTU pointmap | Accuracy (mean, lower better): 1.297 -> 1.011 | Improved |
| add MoE layers (ablation) | MoRE (VGGT-based) · NYUv2 depth | Abs Rel: 0.054 -> 0.051 | Improved |
| add MoE layers (ablation) | MoRE (VGGT-based) · RealEstate10K | AUC@30: 85.14 -> 86.28 | Improved |
| add tailored losses incl. confidence-based depth refinement (without MoE) (ablation) | VGGT baseline · RealEstate10K | AUC@30: 77.62 -> 85.14 | Improved |
| add tailored losses incl. confidence-based depth refinement (without MoE) (ablation) | VGGT baseline · DTU pointmap | Completion (lower better): 1.896 -> 1.625 | Improved |
| add tailored losses (without MoE) (ablation) | VGGT baseline · RealEstate10K | RRA@30: 99.97 -> 99.94 | No clear effect |
| MoRE vs Pi3 on ETH3D pointmap accuracy | MoRE · ETH3D pointmap | Accuracy (mean): 0.194 (Pi3) vs 0.234 (MoRE) | Hurt |
| MoRE vs monocular models on Sintel depth | MoRE · Sintel depth | Abs Rel: 0.273 (MoGe) vs 0.274 | No clear effect |
| add dense semantic (DINOv2) feature fusion for normal head (ablation) | MoRE · Normal estimation | qualitative sharpness | Improved |
| MoRE vs Pi3 on Co3Dv2 camera pose | MoRE · Co3Dv2 | AUC@30: 88.42 vs 88.41 (Pi3), 88.59 (VGGT) | No clear effect |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 2 on most shared results.
- VGGT — VGGT: Visual Geometry Grounded Transformer (CVPR 2025) · better on 7 of 8 shared results
- MoGe — MoGe: Unlocking Accurate Monocular Geometry Estimation for Open-Domain Images with Optimal Training Supervision (CVPR 2025) · better on 0 of 1 shared result
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
- Point-MoE: Large-Scale Multi-Dataset Training with Mixture-of-Experts for 3D Semantic Segmentation (ICLR 2026 · similarity 0.87)
- Multilinear Mixture of Experts: Scalable Expert Specialization through Factorization (NeurIPS 2024 · similarity 0.86)
- Real3D: Towards Scaling Large Reconstruction Models with Real Images (ICCV 2025 · similarity 0.86)
- MoE Jetpack: From Dense Checkpoints to Adaptive Mixture of Experts for Vision Tasks (NeurIPS 2024 · similarity 0.86)
- Robust Geometry-Preserving Depth Estimation Using Differentiable Rendering (ICCV 2023 · similarity 0.86)
- Dens3R: A Foundation Model for 3D Geometry Prediction (ICLR 2026 · similarity 0.86)
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.
