

Paper proposing EFANNA, an extremely fast approximate nearest neighbor search algorithm based on kNN graph construction. The method introduces an efficient approximate kNN graph building approach and a search algorithm that achieves state-of-the-art query performance.
EFANNA is an extremely fast approximate nearest neighbor search algorithm based on kNN graphs, proposed by Cong Fu and colleagues in 2016.
EFANNA's graph construction techniques influenced subsequent graph-based methods and contributed to the development of more efficient ANNS algorithms. The work was extended in several follow-up publications by the same research group.
Loading more......