Research map / 3D vision / Face and shape
Nautilus: Locality-aware Autoencoder for Scalable Mesh Generation
Yuxuan Wang, Xuanyu Yi, Haohan Weng, Qingshan Xu, Xiaokang Wei, Xianghui Yang, Chunchao Guo, Long Chen, Hanwang 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 |
|---|---|---|---|
| add Nautilus Shell representation to coordinate compression (ablation) | Nautilus tokenization · Quantitative ablation | Chamfer distance: 0.151 -> 0.100 | Improved |
| add Nautilus Shell representation to coordinate compression (ablation) | Nautilus tokenization · Quantitative ablation | Hausdorff distance: 0.355 -> 0.213 | Improved |
| add coordinate compression to Nautilus Shell representation (ablation) | Nautilus tokenization · Quantitative ablation | Chamfer distance: 0.100 -> 0.092 | Improved |
| add coordinate compression to shell representation (ablation) | Nautilus tokenization · Tokenization ablation | Compression ratio: 0.484 -> 0.275 | Improved |
| add Nautilus Shell representation to coordinate compression (ablation) | Nautilus tokenization · Tokenization ablation | Compression ratio: 0.549 -> 0.275 | Improved |
| add local point encoder (PointConv) to the dual-stream conditioner (ablation) | Nautilus · Quantitative ablation | Chamfer distance: 0.092 -> 0.087 | Improved |
| add local point encoder to the conditioner (ablation) | Nautilus · Quantitative ablation | Hausdorff distance: 0.186 -> 0.176 | Improved |
| use Nautilus tokenization instead of AMT (ablation) | mesh tokenization · Tokenization comparison | Local ratio: 0.378 -> 0.554 | Improved |
| use Nautilus tokenization instead of EdgeRunner (ablation) | mesh tokenization · Tokenization comparison | Compression ratio: 0.474 -> 0.275 | Improved |
| use Nautilus instead of MeshAnythingV2 (ablation) | point-cloud mesh generation · Topological quality | Manifold meshes: 29.0% -> 83.6% | Improved |
Later papers that compared with this method
1 later accepted paper reports Nautilus as a baseline in their results tables; 1 of them report a better number on most of the results they share with it.
- MeshWeaver: Sparse-Voxel-Guided Surface Weaving for Autoregressive Mesh Generation (CVPR 2026 · better on 1 of 1 shared result)
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.
- MeshAnything — MeshAnything: Artist-Created Mesh Generation with Autoregressive Transformers (ICLR 2025) · better on 14 of 14 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
- ARMesh: Autoregressive Mesh Generation via Next-Level-of-Detail Prediction (NeurIPS 2025 · similarity 0.90)
- EdgeRunner: Auto-regressive Auto-encoder for Artistic Mesh Generation (ICLR 2025 · similarity 0.89)
- MeshAnything: Artist-Created Mesh Generation with Autoregressive Transformers (ICLR 2025 · similarity 0.88)
- MeshGPT: Generating Triangle Meshes with Decoder-Only Transformers (CVPR 2024 · similarity 0.87)
- TreeMeshGPT: Artistic Mesh Generation with Autoregressive Tree Sequencing (CVPR 2025 · similarity 0.86)
- GALA: Geometry-Aware Local Adaptive Grids for Detailed 3D Generation (ICLR 2025 · 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.
