Research map / 3D vision / Face and shape
HoloTetSphere: Unified TetSphere Mesh Reconstruction for Physical Simulations
Yaqiao Dai, Renjiao Yi, Zhirui Gao, Wei Chen, Kai Xu, Chenyang Zhu
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 continuous vertex-shared opacity field for pruning (ablation) | HoloTetSphere · Combined 3D dataset (Thingi10k, DeepFashion3D, Objaverse, GSO) | number of components: 133 -> 1 | Improved |
| Add continuous vertex-shared opacity field (ablation) | HoloTetSphere · Combined 3D dataset (Thingi10k, DeepFashion3D, Objaverse, GSO) | normal consistency: 0.597 -> 0.819 | Improved |
| Use two-stage HC-Laplacian instead of single-stage smoothing (ablation) | HoloTetSphere · Combined 3D dataset (Thingi10k, DeepFashion3D, Objaverse, GSO) | normal consistency: 0.784 -> 0.819 | Improved |
| Use weighted bi-harmonic instead of uniform energy (ablation) | HoloTetSphere · Combined 3D dataset (Thingi10k, DeepFashion3D, Objaverse, GSO) | edge precision: 0.178 -> 0.379 | Improved |
| Use weighted bi-harmonic instead of uniform energy (ablation) | HoloTetSphere · Combined 3D dataset (Thingi10k, DeepFashion3D, Objaverse, GSO) | Volume IoU / F1: IoU 0.645 -> 0.620; F1 0.444 -> 0.428 | Hurt |
| Add two-stage HC smoothing (ablation) | HoloTetSphere · Combined 3D dataset (Thingi10k, DeepFashion3D, Objaverse, GSO) rendering | PSNR: 39.3252 -> 40.9509 | Improved |
| Use topology-adaptive pruning instead of TetSphere homeomorphic deformation | TetSphere · Combined 3D dataset (Thingi10k, DeepFashion3D, Objaverse, GSO) | single-component rate: 0.0% -> 96.7% | Improved |
| Use topology-adaptive pruning instead of TetSphere | TetSphere · Combined 3D dataset (Thingi10k, DeepFashion3D, Objaverse, GSO) | inverted ratio: 0.057 -> 0.017 | Improved |
| Use pruning-based approach vs TetSphere | TetSphere · Combined 3D dataset (Thingi10k, DeepFashion3D, Objaverse, GSO) | manifoldness rate: 100.0% -> 96.7% | Hurt |
| Replace TetSphere with HoloTetSphere | TetSphere · Combined 3D dataset (Thingi10k, DeepFashion3D, Objaverse, GSO) | Chamfer: 0.0173 -> 0.0102 | Improved |
| Joint tetrahedra-Gaussian optimization vs NeuS2 on DTU | NeuS2 · DTU | PSNR: 32.14 -> 35.36 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 2 of 2 on most shared results.
- NeuS2 — NeuS2: Fast Learning of Neural Implicit Surfaces for Multi-view Reconstruction (ICCV 2023) · better on 3 of 5 shared results
- DMesh — DMesh: A Differentiable Mesh Representation (NeurIPS 2024) · better on 3 of 3 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
- TetSphere Splatting: Representing High-Quality Geometry with Lagrangian Volumetric Meshes (ICLR 2025 · similarity 0.91)
- Tetrahedron Splatting for 3D Generation (NeurIPS 2024 · similarity 0.89)
- Radiance Meshes for Volumetric Reconstruction (CVPR 2026 · similarity 0.86)
- Ghost on the Shell: An Expressive Representation of General 3D Shapes (ICLR 2024 · similarity 0.84)
- SILSA: Sliding-Window Slice Latents for Topology-Preserving High-Resolution 3D Generation (NeurIPS 2026 · similarity 0.83)
- Geometry-Centered 3D Latent World Models for Growing Surfaces (NeurIPS 2026 · similarity 0.83)
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.
