Research map / Point clouds / Shape and surface completion
PointAE: Point Auto-Encoder for 3D Statistical Shape and Texture Modelling
Hang Dai, Ling Shao
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 skip connection (ablation) | PointAE · Headspace ablation (L=128) | Normalized mean error (NME): 0.0612 -> 0.0271 | Improved |
| add skip connection (ablation) | PointAE · Headspace ablation (L=1024) | Normalized mean error (NME): 0.0501 -> 0.0196 | Improved |
| add attention block (ablation) | PointAE · Headspace ablation (L=128) | Normalized mean error (NME): 0.0612 -> 0.0537 | Improved |
| add attention block (ablation) | PointAE · Headspace ablation (L=1024) | Normalized mean error (NME): 0.0501 -> 0.0438 | Improved |
| add attention block to model with skip connection (ablation) | PointAE · Headspace ablation (L=128) | Normalized mean error (NME): 0.0271 -> 0.0202 | Improved |
| add skip connection to model with attention (ablation) | PointAE · Headspace ablation (L=128) | Normalized mean error (NME): 0.0537 -> 0.0202 | Improved |
| add correspondence refinement and bootstrapping (ablation) | PointAE · Headspace ablation (L=128) | Normalized mean error (NME): 0.0202 -> 0.0168 | Improved |
| add correspondence refinement and bootstrapping (ablation) | PointAE · Headspace ablation (L=1024) | Normalized mean error (NME): 0.0107 -> 0.0083 | Improved |
| use PointAE instead of Compositional VAE | PointAE · BU3DFE (L=128) | Normalized mean error (NME): 0.0222 -> 0.0178 | Improved |
| use PointAE instead of PCA | PointAE · BU3DFE (L=1024) | Normalized mean error (NME): 0.0187 -> 0.0074 | Improved |
| use PointAE instead of Compositional VAE | PointAE · Headspace (L=1024) | Normalized mean error (NME): 0.0103 -> 0.0083 | Improved |
| use PointAE instead of PCA | PointAE · Headspace (L=128) | Normalized mean error (NME): 0.0334 -> 0.0168 | Improved |
Closest papers by meaning
- Learning 3D Dense Correspondence via Canonical Point Autoencoder (NeurIPS 2021 · similarity 0.86)
- Point-M2AE: Multi-scale Masked Autoencoders for Hierarchical Point Cloud Pre-training (NeurIPS 2022 · similarity 0.86)
- Towards Compact 3D Representations via Point Feature Enhancement Masked Autoencoders (AAAI 2024 · similarity 0.84)
- ShapeEmbed: a self-supervised learning framework for 2D contour quantification (NeurIPS 2025 · similarity 0.84)
- Dora: Sampling and Benchmarking for 3D Shape Variational Auto-Encoders (CVPR 2025 · similarity 0.84)
- Adaptive Siamese Masked Autoencoder with Global Optimization for Unsupervised Point Cloud Shape Correspondence (AAAI 2025 · similarity 0.84)
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.
