Research map / Learning theory and algorithms / Optimal transport
Near-Linear Time Algorithm for the Chamfer Distance
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 LSH-based importance sampling instead of uniform sampling | Chamfer distance estimation · ShapeNet | samples for approximately 2% relative error: 100 vs 500 samples | Helped |
| Use LSH-based importance sampling instead of uniform sampling | Chamfer distance estimation · Federalist Papers text embeddings | samples for 2% relative error: 20 vs at least 450 samples | Helped |
| Use LSH-based importance sampling instead of uniform sampling | Chamfer distance estimation · Outliers | relative error | Helped |
| Use 3 grid levels for crude nearest-neighbor estimates (ablation) | Chamfer-Estimate · ShapeNet | crude NN estimate factor from true nearest-neighbor distance: 3 grid levels; within a factor of 2 | Helped |
| Decrease the number of LSH structures or the window size (ablation) | Chamfer-Estimate on DEEP1B and Microsoft-Turing · DEEP1B and Microsoft-Turing | importance-sampling sample complexity trade-off | Hurt |
| Increase LSH structures and window size to 64 \times 10^6 (ablation) | Chamfer-Estimate on DEEP1B and Microsoft-Turing · DEEP1B and Microsoft-Turing | sample complexity trade-off: best parameter choice (both 64 \times 10^6) | Helped |
| Use 16 LSH structures and window size 2 \cdot 10^5 instead of uniform sampling | Chamfer distance estimation · DEEP1B | runtime for 1% relative error: approximately 1785 seconds vs >2200 seconds | Helped |
| Use approximate Chamfer distances to prune candidates before exact evaluation | ShapeNet nearest point-cloud retrieval · ShapeNet | recall@30: >98% | Helped |
| Use importance sampling instead of uniform sampling | Chamfer distance estimation · ShapeNet | runtime: 50% faster | Helped |
Closest papers by meaning
- Fully Dynamic Algorithms for Chamfer Distance (NeurIPS 2025 · similarity 0.93)
- CMG-Net: Robust Normal Estimation for Point Clouds via Chamfer Normal Distance and Multi-Scale Geometry (AAAI 2024 · similarity 0.84)
- Approximating Metric Magnitude of Point Sets (AAAI 2025 · similarity 0.83)
- Beyond Chamfer Distance: Granular Order-aware Evaluation Metric For Online Mapping (NeurIPS 2026 · similarity 0.82)
- An in depth look at the Procrustes-Wasserstein distance: properties and barycenters (ICML 2025 · similarity 0.82)
- Euclidean distance compression via deep random features (NeurIPS 2024 · similarity 0.82)
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.