Jan 2, 2019 · In this paper we propose a novel hierarchical inverted index structure generated by vector and line quantization methods.
A novel hierarchical inverted indexing structure based on vector and line quantization for accurate billion-scale approximate nearest neighbor search. •. A ...
A novel GPU-friendly search structure based on nearest neighbor graphs and information propagation on graphs is proposed that significantly surpasses the ...
Our evaluation shows that VLQ-ADC significantly outperforms the state-of-the-art GPU- and CPU-based systems in terms of both accuracy and search speed.
In this paper we propose a novel hierarchical inverted index structure generated by vector and line quantization methods. Our quantization method improves both ...
Bibliographic details on Vector and line quantization for billion-scale similarity search on GPUs.
Feb 28, 2017 · The most popular vector compression methods can be classified into either binary codes [18, 22], or quantization methods [25, 37]. Both have the ...
Aug 11, 2023 · To query nq vectors, each replica handles a fraction nq/R of the queries, joining the results back together on a single GPU or in CPU memory.
Missing: line | Show results with:line
structing the vectors. Our paper focuses on methods based on product quanti- zation (PQ) codes, as these were shown to be more effective
May 25, 2024 · The paper presents a comprehensive approach to efficient similarity search on GPUs, covering both exact and approximate search methods.
Missing: line | Show results with:line