We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:

Download:

Current browse context:

cs.DS

Change to browse by:

References & Citations

DBLP - CS Bibliography

Bookmark

(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Computer Science > Data Structures and Algorithms

Title: Falconn++: A Locality-sensitive Filtering Approach for Approximate Nearest Neighbor Search

Authors: Ninh Pham, Tao Liu
Abstract: We present Falconn++, a novel locality-sensitive filtering approach for approximate nearest neighbor search on angular distance. Falconn++ can filter out potential far away points in any hash bucket \textit{before} querying, which results in higher quality candidates compared to other hashing-based solutions. Theoretically, Falconn++ asymptotically achieves lower query time complexity than Falconn, an optimal locality-sensitive hashing scheme on angular distance. Empirically, Falconn++ achieves higher recall-speed tradeoffs than Falconn on many real-world data sets. Falconn++ is also competitive with HNSW, an efficient representative of graph-based solutions on high search recall regimes.
Comments: To appear in NeurIPS 2022
Subjects: Data Structures and Algorithms (cs.DS); Computer Vision and Pattern Recognition (cs.CV)
Cite as: arXiv:2206.01382 [cs.DS]
  (or arXiv:2206.01382v3 [cs.DS] for this version)

Submission history

From: Ninh Pham [view email]
[v1] Fri, 3 Jun 2022 04:02:02 GMT (541kb)
[v2] Wed, 14 Sep 2022 23:01:38 GMT (541kb)
[v3] Sat, 22 Oct 2022 08:45:28 GMT (826kb)

Link back to: arXiv, form interface, contact.