Research map / Point clouds / Point cloud classification
Towards a Persistence Diagram that is Robust to Noise and Varied Densities
Hang Zhang, Kaifeng Zhang, Kai Ming Ting, Ye 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 |
|---|---|---|---|
| Replace distance filter with Λ-filter (data-dependent Voronoi kernel) | persistence diagram construction · immune cell t-SNE visualization | kNN classification accuracy: 0.89 vs 0.8 for DTM | Improved |
| Replace distance filter with Λ-filter | persistence diagram features for SVM · bone script classification | SVM accuracy | Improved |
| Use Λ-filter instead of Gaussian-kernel (GK-filter) data-independent filter | persistence diagram under noise · Cassini noise robustness | bottleneck distance D(X) vs D(X∪Xn) at γ=0.025: 0.03 vs 0.17 | Improved |
| Use Λ-filter instead of DTM | persistence diagram under noise · Cassini noise robustness | bottleneck distance at γ=0.025: 0.03 vs 0.08 | Improved |
| Use Λ-filter instead of DTM/Rips under stretching and shrinking | persistence diagram under varied densities · Stretching/shrinking of Cassini-type point cloud | bottleneck distance between D(X) and D(X_r) | Improved |
| Use CkNN normalization instead of DTM/Rips under stretching | persistence diagram under varied densities · noise-free stretching/shrinking | bottleneck distance | Improved |
| Add one noise point; CkNN loses robustness to shrinking | CkNN · stretching/shrinking with one noise point | bottleneck distance | Hurt |
| Use Λ-filter with finite η instead of η=∞ in SVM bone script classification (ablation) | Λ-filter · bone script classification | SVM accuracy | Improved |
| Use DTM instead of Λ-filter on mixed noise and varied-density point cloud | DTM · two rings with noise and varied density | number of persistent rings | Hurt |
| Use Λ-filter instead of GK-filter in noise sensitivity test | persistence diagram under noise · Cassini noise test | correct number of valid rings | Improved |
Closest papers by meaning
- Persistence weighted Gaussian kernel for topological data analysis (ICML 2016 · similarity 0.92)
- Kernel Method for Persistence Diagrams via Kernel Embedding and Weight Factor (JMLR 2018 · similarity 0.91)
- Topo-Geometric Analysis of Variability in Point Clouds Using Persistence Landscapes (TPAMI 2024 · similarity 0.90)
- Point-Level Topological Representation Learning on Point Clouds (ICML 2025 · similarity 0.88)
- A Class of Topological Pseudodistances for Fast Comparison of Persistence Diagrams (AAAI 2024 · similarity 0.87)
- Diffeomorphic interpolation for efficient persistence-based topological optimization (NeurIPS 2024 · similarity 0.87)
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.
