Filters








28,470 Hits in 4.4 sec

Trading Quality for Time with Nearest-Neighbor Search [chapter]

Roger Weber, Klemens Böhm
2000 Lecture Notes in Computer Science  
In this article, we investigate approximate evaluation techniques based on the VA-File for Nearest-Neighbor Search (NN-Search). The VA-File contains approximations of feature points.  ...  Experiments show that these techniques have the desired effect: for instance, when allowing for a small but specific reduction of result quality, we observed a speedup of 7 in 50-NN search.  ...  Schek and Martin Breunig for helpful comments.  ... 
doi:10.1007/3-540-46439-5_2 fatcat:s42rkrekofc7nctoy7dwd6bbwi

GGNN: Graph-based GPU Nearest Neighbor Search [article]

Fabian Groh, Lukas Ruppert, Patrick Wieschollek, Hendrik P.A. Lensch
2021 arXiv   pre-print
Approximate nearest neighbor (ANN) search in high dimensions is an integral part of several computer vision systems and gains importance in deep learning with explicit memory representations.  ...  In this paper, we propose a novel search structure based on nearest neighbor graphs and information propagation on graphs.  ...  In Figure 5 , the trade off between build time and query time for a fixed recall rate is visualized.  ... 
arXiv:1912.01059v3 fatcat:zbewjskznrhexkvt2zc6vacnqy

Efficient approximate nearest neighbor search with integrated binary codes

Wei Zhang, Ke Gao, Yongdong Zhang, Jintao Li
2011 Proceedings of the 19th ACM international conference on Multimedia - MM '11  
The difficulty of exact nearest neighbor search has led to approximate solutions that sacrifice precision for efficiency.  ...  Nearest neighbor search in Euclidean space is a fundamental problem in multimedia retrieval.  ...  For these binary code methods, searching nearest neighbors in Euclidean space is approximated by searching similar neighbors in terms of Hamming distances between codes.  ... 
doi:10.1145/2072298.2071971 dblp:conf/mm/ZhangGZL11 fatcat:ulalbkj5rfd5hmcitf3g2wtfny

Scalable Nearest Neighbor Search based on kNN Graph [article]

Wan-Lei Zhao, Jie Yang, Cheng-Hao Deng
2017 arXiv   pre-print
Nearest neighbor search is known as a challenging issue that has been studied for several decades.  ...  We show that our method achieves the best trade-off between search quality, efficiency and memory complexity.  ...  With the support of kNN graph, nearest neighbor search is conducted by hill-climbing strategy [4] .  ... 
arXiv:1701.08475v2 fatcat:awotobncyfhizizl64niebs2lq

An efficient computation-constrained block-based motion estimation algorithm for low bit rate video coding

M. Gallant, G. Cote, F. Kossentini
1999 IEEE Transactions on Image Processing  
A reliable predictor determines the search origin, localizing the search process. An e cient search pattern exploits structural constraints within the motion eld.  ...  The resulting low bit rate video encoder yields essentially the same levels of rate-distortion performance and subjective quality a c hieved by the UBC H.263+ video coding reference software.  ...  Recall that the high-quality pro le illustrated in Figure 6 is the pro le used to obtain the results for the nearest-neighbors search in Figures 4 and 5 .  ... 
doi:10.1109/83.806627 pmid:18267458 fatcat:5dgffa5bmzhrbbmdrvco6nzp7a

Similarity search on neighbor's graphs with automatic Pareto optimal performance and minimum expected quality setups based on hyperparameter optimization [article]

Eric S. Tellez, Guillermo Ruiz
2022 arXiv   pre-print
This manuscript introduces an autotuned algorithm for searching nearest neighbors based on neighbor graphs and optimization metaheuristics to produce Pareto-optimal searches for quality and search speed  ...  Our approach is described and benchmarked with other state-of-the-art similarity search methods, showing convenience and competitiveness.  ...  The first two aim to find the best trade between quality and search speed, and the second one is the best trade ensuring a given minimum quality.  ... 
arXiv:2201.07917v1 fatcat:a5vhouck4jfxhjud2ojwno26yy

Anytime k-nearest neighbor search for database applications

Weijia Xu, Daniel Miranker, Rui Mao, Smriti Ramakrishnan
2008 2008 IEEE 24th International Conference on Data Engineering Workshop  
Given unbounded time the algorithm terminates with an exact solution. Approximate solutions to k-nearest neighbor search provide much needed speed improvement to hard nearest-neighbor problems.  ...  In support of these applications, we explored a search strategy for knearest neighbor searches with MVP-trees that greedily identifies k answers and then improves the answer set monotonically.  ...  terminated k nearest neighbor search is also two to four times faster than a completed k nearest neighbor search with slightly worse accuracy.  ... 
doi:10.1109/icdew.2008.4498354 dblp:conf/icde/XuMMR08 fatcat:vppvtvfoinghre3nfmx6phqyjq

Anytime K-Nearest Neighbor Search for Database Applications

Weijia Xu, Daniel P. Miranker, Rui Mao, Smriti Ramakrishnan
2008 First International Workshop on Similarity Search and Applications (sisap 2008)  
Given unbounded time the algorithm terminates with an exact solution. Approximate solutions to k-nearest neighbor search provide much needed speed improvement to hard nearest-neighbor problems.  ...  In support of these applications, we explored a search strategy for knearest neighbor searches with MVP-trees that greedily identifies k answers and then improves the answer set monotonically.  ...  terminated k nearest neighbor search is also two to four times faster than a completed k nearest neighbor search with slightly worse accuracy.  ... 
doi:10.1109/sisap.2008.11 dblp:conf/sisap/XuMMR08 fatcat:nxr4l7iv7bbupgj5gw6olk6xyu

Speeding up the Xbox recommender system using a euclidean transformation for inner-product spaces

Yoram Bachrach, Yehuda Finkelstein, Ran Gilad-Bachrach, Liran Katzir, Noam Koenigstein, Nir Nice, Ulrich Paquet
2014 Proceedings of the 8th ACM Conference on Recommender systems - RecSys '14  
We propose a novel order preserving transformation, mapping the maximum inner product search problem to Euclidean space nearest neighbor search problem.  ...  We evaluate our techniques on two large-scale recommendation datasets, Xbox Movies and Yahoo Music, and show that this technique allows trading off a slight degradation in the recommendation quality for  ...  In construction time, nodes are split along one coordinate. At query time, one can search of all points in a rectangular box and nearest neighbors efficiently.  ... 
doi:10.1145/2645710.2645741 dblp:conf/recsys/BachrachFGKKNP14 fatcat:qwptdh3mefcmjkvpe4y34ca2jm

Fast k-means based on KNN Graph [article]

Cheng-Hao Deng, Wan-Lei Zhao
2017 arXiv   pre-print
Comparing with existing fast k-means variants, the proposed algorithm achieves hundreds to thousands times speed-up while maintaining high clustering quality.  ...  In the proposal, k-means is supported by an approximate k-nearest neighbors graph. In the k-means iteration, each data sample is only compared to clusters that its nearest neighbors reside.  ...  on the approximate nearest neighbor search (ANNS) tasks.  ... 
arXiv:1705.01813v1 fatcat:ypnw62f7q5d4rnigdkyycusuta

Barnes-Hut-SNE [article]

Laurens van der Maaten
2013 arXiv   pre-print
The paper presents an O(N log N)-implementation of t-SNE -- an embedding technique that is commonly used for the visualization of high-dimensional data in scatter plots and that normally runs in O(N^2)  ...  Our experiments show that the new algorithm, called Barnes-Hut-SNE, leads to substantial computational advantages over standard t-SNE, and that it makes it possible to learn embeddings of data sets with  ...  The author thanks Geoffrey Hinton for many helpful discussions, and two anonymous reviewers for their helpful comments.  ... 
arXiv:1301.3342v2 fatcat:jgu6ymgpjnhghptcgd4s53i6im

Large-Scale Visual Search with Binary Distributed Graph at Alibaba

Kang Zhao, Pan Pan, Yun Zheng, Yanhao Zhang, Changxu Wang, Yingya Zhang, Yinghui Xu, Rong Jin
2019 Proceedings of the 28th ACM International Conference on Information and Knowledge Management - CIKM '19  
Graph-based approximate nearest neighbor search has attracted more and more attentions due to its online search advantages.  ...  For a deployed visual search system with several billions of online images in total, building a billion-scale offline graph in hours is essential, which is almost unachievable by most existing methods.  ...  In order to get high recall for query locating anywhere, fast approximate nearest neighbor graph (FANNG) improves the search algorithm with backtracking.  ... 
doi:10.1145/3357384.3357834 dblp:conf/cikm/ZhaoPZZWZXJ19 fatcat:yvqezafoz5appoec2a3eelwuqu

Learning Space Partitions for Nearest Neighbor Search [article]

Yihe Dong and Piotr Indyk and Ilya Razenshteyn and Tal Wagner
2020 arXiv   pre-print
Space partitions of ℝ^d underlie a vast and important class of fast nearest neighbor search (NNS) algorithms.  ...  Inspired by recent theoretical work on NNS for general metric spaces [Andoni, Naor, Nikolov, Razenshteyn, Waingarten STOC 2018, FOCS 2018], we develop a new framework for building space partitions reducing  ...  Introduction The Nearest Neighbor Search (NNS) problem is defined as follows.  ... 
arXiv:1901.08544v4 fatcat:gyt2wqp6uvamnjckbnggqilszi

Balancing clusters to reduce response time variability in large scale image search [article]

Romain Tavenard , Hervé Jégou
2010 arXiv   pre-print
Many algorithms for approximate nearest neighbor search in high-dimensional spaces partition the data into clusters.  ...  Experiments with a large scale collection of image descriptors show that our algorithm significantly reduces the variance of response times without seriously impacting the search quality.  ...  Metrics: Selectivity and Recall All approximate nearest-neighbor search methods try to find the best trade-off between result quality and retrieval time.  ... 
arXiv:1009.4739v1 fatcat:hdod6pwlgbbwnpwhfvhiel45vm

A novel index structure for large scale image descriptor search

Jiangbo Yuan, Xiuwen Liu
2012 2012 19th IEEE International Conference on Image Processing  
This paper presents a k-means based algorithm for approximate nearest neighbor search.  ...  Index Termsk-means, approximate nearest neighbor search, multi-assignment, pruning strategies, image descriptor indexing  ...  The scalability and accuracy of traditional nearest neighbor search algorithms are severely limited by the curses of dimensionality for data in high dimensional.  ... 
doi:10.1109/icip.2012.6467265 dblp:conf/icip/YuanL12 fatcat:u2qdtyljunbdxpbie6laa42feu
« Previous Showing results 1 — 15 out of 28,470 results