Research map / Point clouds / Shape and surface completion
Learning to Tessellate: Point Cloud Generation via Recursive Spectral Partitioning
Monan Sun, Bangzhen Liu, Huaidong Zhang, Shengfeng He
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 cascaded generation (ablation) | PointNSP-style baseline · ShapeNetV2 Car ablation | 1-NNA CD (%): 53.66 -> 53.02 | Improved |
| add cascaded generation (ablation) | PointNSP-style baseline · ShapeNetV2 Car ablation | Std MMD CD: 0.357 -> 0.343 (x10^-3) | Improved |
| add balanced KNN binary-tree structural prior (ablation) | cascaded baseline (V1) · ShapeNetV2 Car ablation | 1-NNA CD (%): 53.02 -> 52.75 | Improved |
| add balanced KNN binary-tree structural prior (ablation) | cascaded baseline (V1) · ShapeNetV2 Car ablation | Std MMD EMD: 0.923 -> 0.713 (x10^-3) | Improved |
| replace balanced KNN tree with topology-aware partitioning autoencoder (recursive spectral partitioning) (ablation) | V2 · ShapeNetV2 Car ablation | 1-NNA CD (%): 52.75 -> 51.89 | Improved |
| replace balanced KNN tree with topology-aware partitioning autoencoder (ablation) | V2 · ShapeNetV2 Car ablation | 1-NNA EMD (%): 53.65 -> 53.01 | Improved |
| add geometry-calibrated positional encoding (ablation) | V3 · ShapeNetV2 Car ablation | 1-NNA CD (%): 51.89 -> 51.47 | Improved |
| add geometry-calibrated positional encoding (ablation) | V3 · ShapeNetV2 Car ablation | Std MMD EMD: 0.636 -> 0.506 (x10^-3) | Improved |
| use PointRSP instead of PointNSP | PointNSP · ShapeNetV2 single-category (mean) | 1-NNA CD (%): 60.88 -> 58.67 | Improved |
| use PointRSP instead of PointNSP | PointNSP · ShapeNetV2 single-category (mean) | 1-NNA EMD (%): 57.67 -> 56.86 | Improved |
| use PointRSP instead of PointNSP in multi-class generation | PointNSP · ShapeNetV2 multi-class Chair | 1-NNA EMD (%): 55.10 -> 56.07 | Hurt |
| use PointRSP instead of PointNSP in multi-class generation | PointNSP · ShapeNetV2 multi-class Car | 1-NNA CD (%): 69.97 -> 58.39 | 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.
- TIGER — TIGER: Time-Varying Denoising Model for 3D Point Cloud Generation with Diffusion Process (CVPR 2024) · better on 2 of 2 shared results
- LION — LION: Latent Point Diffusion Models for 3D Shape Generation (NeurIPS 2022) · better on 2 of 2 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
- PointNSP: Autoregressive 3D Point Cloud Generation with Next-Scale Level-of-Detail Prediction (CVPR 2026 · similarity 0.93)
- Quartet of Diffusions: Structure-Aware Point Cloud Generation through Part and Symmetry Guidance (ICLR 2026 · similarity 0.90)
- Autoregressive 3D Shape Generation via Canonical Mapping (ECCV 2022 · similarity 0.90)
- Accelerated Generative Models for 3D Point Cloud Data (CVPR 2016 · similarity 0.89)
- TopNet: Structural Point Cloud Decoder (CVPR 2019 · similarity 0.89)
- PointInfinity: Resolution-Invariant Point Diffusion Models (CVPR 2024 · 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.
