Research map / Graph learning / Graph algorithms and clustering
FANNG: Fast Approximate Nearest Neighbour Graphs
Check 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 the thresholded occlusion rule with τ/τmax = 0.38 instead of 0.00 (ablation) | FANNG · 100K SIFT | recall: 0.7098 -> 0.9984 | Helped |
| Increase τ/τmax from 0.00 to 0.38 (ablation) | FANNG · 100K SIFT | average cost per query: 110.52 -> 360.05 | Hurt |
| Increase τ/τmax from 0.00 to 0.50 (ablation) | FANNG · 100K SIFT | recall: 0.7098 -> 0.9999 | Helped |
| Increase τ/τmax from 0.00 to 0.75 (ablation) | FANNG · 100K SIFT | average cost per query: 110.52 -> 1315.5 | Hurt |
| Increase τ/τmax from 0.00 to 1.00 (ablation) | FANNG · 100K SIFT | average degree: 24.329 -> 2182.9 | Hurt |
| Truncate each vertex's outgoing edges to at most 25 (ablation) | FANNG ideal graph · SIFT | search efficiency | Helped |
| Use occlusion-pruned edges instead of a plain k-nearest-neighbour graph (ablation) | FANNG · SIFT | computational efficiency | Helped |
| Use GPU FANNG search instead of linear search at 90% recall | SIFT nearest-neighbour queries · 1M SIFT | query speed: approximately 500 times faster | Helped |
| Use GPU FANNG search instead of linear search at 90% recall | SIFT nearest-neighbour queries · 5M SIFT | query speed: approximately 1400 times faster | Helped |
Closest papers by meaning
- Graph-Based Algorithms for Diverse Similarity Search (ICML 2025 · similarity 0.89)
- MCGI: Manifold-Consistent Graph Indexing for Billion-Scale Disk-Resident Vector Search (NeurIPS 2026 · similarity 0.87)
- CSPG: Crossing Sparse Proximity Graphs for Approximate Nearest Neighbor Search (NeurIPS 2024 · similarity 0.86)
- Distance Adaptive Beam Search for Provably Accurate Graph-Based Nearest Neighbor Search (NeurIPS 2025 · similarity 0.86)
- LotusFilter: Fast Diverse Nearest Neighbor Search via a Learned Cutoff Table (CVPR 2025 · similarity 0.86)
- Approximate Nearest Neighbor Search with Window Filters (ICML 2024 · similarity 0.86)
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.