HEXA: A Disjoint-Subgraph-Based Indexing Framework for Approximate Nearest Neighbor Search at Billion Scale
Summary: HEXA abandons global graph connectivity, independently building GPU-refined proximity subgraphs over disjoint clusters with two-level routing and budget-adaptive search. At billion scale, it cuts construction to ~1 hour and delivers up to 14.7× higher throughput at equal recall. (summarized by gpt-5.6-luna on Aug 28 2026)
Incoming Non-self Citations Over Time
No non-self incoming citations found for this paper in this database.
Authors
- 1. Yifei Xu (Shanghai Jiao Tong University)
- 2. Yanyan Shen (Shanghai Jiao Tong University)
- 3. Youmin Chen (Shanghai Jiao Tong University)
- 4. Linpeng Huang (Shanghai Jiao Tong University)
BibTeX Citation
@article{xu_vldb26,
title = {{HEXA: A Disjoint-Subgraph-Based Indexing Framework for Approximate Nearest Neighbor Search at Billion Scale}},
author = {Xu, Yifei and Shen, Yanyan and Chen, Youmin and Huang, Linpeng},
journal = {PVLDB},
series = {{VLDB} '26},
volume = {19},
number = {10},
pages = {2922--2935},
doi = {10.14778/3828612.3828642},
url = {https://doi.org/10.14778/3828612.3828642},
year = {2026}
}Incoming Citations (Sorted by Pagerank)
Showing 0 of 0 citing papers.
| Rank | Citing Paper | Year | Venue | Pagerank |
|---|
Previous
Page 1 / 1
Next
Outgoing Citations (Sorted by Pagerank)
Showing 12 of 12 cited papers.
Citations counted here include only citations to other VLDB/SIGMOD/CIDR/PODS papers in this database.
Previous
Page 1 / 1
Next