Computer Science > Information Retrieval
[Submitted on 2 Apr 2022]
Title:Graph-based Approximate NN Search: A Revisit
View PDFAbstract:Nearest neighbor search plays a fundamental role in many disciplines such as multimedia information retrieval, data-mining, and machine learning. The graph-based search approaches show superior performance over other types of approaches in recent studies. In this paper, the graph-based NN search is revisited. We optimize two key components in the approach, namely the search procedure and the graph that supports the search. For the graph construction, a two-stage graph diversification scheme is proposed, which makes a good trade-off between the efficiency and reachability for the search procedure that builds upon it. Moreover, the proposed diversification scheme allows the search procedure to decide dynamically how many nodes should be visited in one node's neighborhood. By this way, the computing power of the devices is fully utilized when the search is carried out under different circumstances. Furthermore, two NN search procedures are designed respectively for small and large batch queries on the GPU. The optimized NN search, when being supported by the two-stage diversified graph, outperforms all the state-of-the-art approaches on both the CPU and the GPU across all the considered large-scale datasets.
References & Citations
Bibliographic and Citation Tools
Bibliographic Explorer (What is the Explorer?)
Litmaps (What is Litmaps?)
scite Smart Citations (What are Smart Citations?)
Code, Data and Media Associated with this Article
CatalyzeX Code Finder for Papers (What is CatalyzeX?)
DagsHub (What is DagsHub?)
Gotit.pub (What is GotitPub?)
Papers with Code (What is Papers with Code?)
ScienceCast (What is ScienceCast?)
Demos
Recommenders and Search Tools
Influence Flower (What are Influence Flowers?)
Connected Papers (What is Connected Papers?)
CORE Recommender (What is CORE?)
arXivLabs: experimental projects with community collaborators
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.