Research map / Learning theory and algorithms / Matrix and tensor methods
Spherical Principal Curves
Jongmin Lee, Jang-Hyun Kim, Hee‐Seok Oh
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 |
|---|---|---|---|
| use exact continuous projection instead of Hauberg's finite-point projection | extrinsic principal curve · Earthquake data on S2 (T=77, q=0.1) | reconstruction error: 4.920 -> 0.463 | Improved |
| use exact continuous projection instead of Hauberg's projection | extrinsic principal curve · Earthquake data on S2 (T=77, q=0.1) | distinct projection points: 9/77 -> 76/77 | Improved |
| use exact continuous projection instead of Hauberg's projection | extrinsic principal curve · Motion capture data on S2 (T=500, q=0.03) | distinct projection points: 119/338 -> 332/338 | Improved |
| use exact projection instead of Hauberg's method | spherical principal curve · Simulated Wave on S2 (T=500, sigma=0.07, q=0.05) | reconstruction error: 2.006 -> 0.535 | Improved |
| use exact projection instead of Hauberg's method | spherical principal curve · Simulated Circle on S2 (T=500, sigma=0.07, q=0.05) | reconstruction error: 0.089 -> 0.088 | No clear effect |
| use proposed curves instead of Hauberg's method on 4-sphere | spherical principal curve · Simulated waveform on S4 (T=300, q=0.03) | reconstruction error: 1.990 -> 0.211 | Improved |
| use intrinsic instead of extrinsic expectation step | spherical principal curve · Earthquake data on S2 (T=77, q=0.2) | reconstruction error: 2.662 -> 4.391 | Hurt |
| use intrinsic instead of extrinsic expectation step | spherical principal curve · Simulated waveform on S4 (T=300, q=0.03) | reconstruction error: 0.211 -> 0.729 | Hurt |
| decrease kernel bandwidth q (e.g. 0.01) (ablation) | spherical principal curve · Earthquake data on S2 | curve smoothness | Hurt |
| increase kernel bandwidth q to 0.1 in waveform data (ablation) | spherical principal curve · Simulated waveform on S2 (T=500) | curve fit | Hurt |
| use only T=10 curve points instead of larger T (ablation) | spherical principal curve · Simulated waveform on S2 (q=0.06) | curve fit | Improved |
Closest papers by meaning
- Principal Curves on Riemannian Manifolds (TPAMI 2016 · similarity 0.90)
- Probabilistic Principal Curves on Riemannian Manifolds (TPAMI 2024 · similarity 0.87)
- Sparse Exact PGA on Riemannian Manifolds (ICCV 2017 · similarity 0.84)
- Gauge Invariant Framework for Shape Analysis of Surfaces (TPAMI 2016 · similarity 0.83)
- Intrinsic Riemannian Cross-covariance for Manifold-valued Random Objects (NeurIPS 2026 · similarity 0.82)
- Finsler Multi-Dimensional Scaling: Manifold Learning for Asymmetric Dimensionality Reduction and Embedding (CVPR 2025 · similarity 0.81)
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.
