Filters








35,192 Hits in 6.5 sec

High-Dimensional Image Indexing Based on Adaptive Partitioning and Vector Approximation [chapter]

Guang-Ho Cha
2002 Visual and Multimedia Information Management  
Recently, we developed the LPC-file (Cha et aI. 2(02) for indexing high-dimensional data based on vector approximation.  ...  The basic idea is to adaptively partition the data space to find subspaces with high-density clusters and to assign more bits to them to increase the discriminatory power of the vector approximation.  ...  We stated and discussed the problems associated with the current vector approximation approach for high-dimensional indexing.  ... 
doi:10.1007/978-0-387-35592-4_7 fatcat:oarpoy2swzgpbglsgukjoxhm6u

An Adaptive Index Structure for High-Dimensional Similarity Search [chapter]

P. Wu, B. S. Manjunath, S. Chandrasekaran
2001 Lecture Notes in Computer Science  
A practical method for creating a high dimensional index structure that adapts to the data distribution and scales well with the database size, is presented.  ...  Typical media descriptors, such as texture features, are high dimensional and are not uniformly distributed in the feature space.  ...  Ackowledgement: This research was in part supported by the following grants/awards: LLNL/ISCR award #0108, NSF-IRI 9704785, NSF Instrumentation #EIA-9986057, NSF Infrastructure NSF#EIA-0080134, and by  ... 
doi:10.1007/3-540-45453-5_10 fatcat:oj746xsqizhrdba56vtkheuyqm

The GC-tree: a high-dimensional index structure for similarity search in image databases

Guang-Ho Cha, Chin-Wan Chung
2002 IEEE transactions on multimedia  
The GC-tree is based on a special subspace partitioning strategy which is optimized for a clustered high-dimensional image dataset.  ...  The basic ideas are threefold: 1) we adaptively partition the data space based on a density function that identifies dense and sparse regions in a data space; 2) we concentrate the partition on the dense  ...  Until now, the multidimensional index structures based on the conventional data partitioning have been defeated by the high dimensionality.  ... 
doi:10.1109/tmm.2002.1017736 fatcat:5l6574iv6fewlff3uox2ptklqu

Fast Similarity Search for High-Dimensional Dataset

Quan Wang, Suya You
2006 Eighth IEEE International Symposium on Multimedia (ISM'06)  
A number of theoretical and experimental studies lead us to pursue a new approach, called Fast Filtering Vector Approximation (FFVA) to tackle the problem.  ...  This paper addresses the challenging problem of rapidly searching and matching high-dimensional features for the applications of multimedia database retrieval and pattern recognition.  ...  Instead of partitioning the input data space hierarchically, the vector approximation methods directly index the objects based on linear and flat structure.  ... 
doi:10.1109/ism.2006.78 dblp:conf/ism/WangY06 fatcat:aitj4tlzwvb6xnno4tfzo6e3za

Partitioned vector quantization: application to lossless compression of hyperspectral images

G. Motta, F. Rizzo, J.A. Storer
2003 2003 International Conference on Multimedia and Expo. ICME '03. Proceedings (Cat. No.03TH8698)  
High dimensional source vectors are first partitioned into two or more subvectors of (possibly) different length and then, each subvector is individually encoded with an appropriate codebook.  ...  This scheme allows practical quantization of high dimensional vectors in which each vector component is allowed to have different alphabet and distribution.  ...  The proposed algorithm is intended for the encoding of source vectors drawn from a high dimensional source on D R .  ... 
doi:10.1109/icme.2003.1220977 dblp:conf/icmcs/MottaRS03 fatcat:jcpn7ho4xrerrg3rg6y6isemeq

An efficient high-dimensional indexing method for content-based retrieval in large image databases

I. Daoudi, K. Idrissi, S.E. Ouatik, A. Baskurt, D. Aboutajdine
2009 Signal processing. Image communication  
Based on Euclidian distance, many of them have been proposed for applications where data vectors are high-dimensional.  ...  In this way, the proposed approach achieves high performances in response time and in precision when dealing with high-dimensional and heterogeneous vectors.  ...  The RA-Blocks outperforms other methods based on the approximation vector approach in very high-dimensional spaces.  ... 
doi:10.1016/j.image.2009.09.001 fatcat:v5ce4yqlw5en5gc5sbwk763hru

Adaptive nearest neighbor search for relevance feedback in large image databases

P. Wu, B. S. Manjunath
2001 Proceedings of the ninth ACM international conference on Multimedia - MULTIMEDIA '01  
The main contribution of this paper is a novel algorithm for adaptive nearest neighbor computations for high dimensional feature vectors and when the number of items in the database is large.  ...  Such nearest neighbor computations are expensive given that typical image features, such as color and texture, are represented in high dimensional spaces.  ...  An approximation based index scheme to support relevance feedback Typical audio-visual descriptors are high dimensional vectors [1, 11] . Their dimensionality range from few tens to a few hundreds.  ... 
doi:10.1145/500156.500157 fatcat:inqd3roqoffupeujup3xp4fcwi

Transactional Support for Visual Instance Search [chapter]

Herwig Lejsek, Friðrik Heiðar Ásmundsson, Björn Þór Jónsson, Laurent Amsaleg
2018 Lecture Notes in Computer Science  
Quantization One line of work is based on indexing data clusters such as the hierarchical k-means decomposition of the data collection: Voronoi cells are created to partition and store the high-dimensional  ...  The NV-tree The NV-tree [14, 15] is a disk-based high-dimensional index, based upon a combination of projections of data points to lines and partitioning of the projected space.  ... 
doi:10.1007/978-3-030-02224-2_6 fatcat:b3h3opoxyndwnetzcui7by2dci

A Survey on Efficient Processing of Similarity Queries over Neural Embeddings [article]

Yifan Wang
2022 arXiv   pre-print
Then we talk about recent approaches on designing the indexes and operators for highly efficient similarity query processing on top of embeddings (or more generally, high dimensional data).  ...  semantics of the raw data, based on which embeddings do show outstanding effectiveness on capturing data similarities, making it one of the most widely used and studied techniques in the state-of-the-art  ...  Traditional similarity/distance based indexes like KD-tree work well on low-dimensional data but perform weakly on high-dimensional data due to the curse of dimensionality.  ... 
arXiv:2204.07922v1 fatcat:u5osyghs6vgppnj5gpnrzhae5y

Dynamicity and Durability in Scalable Visual Instance Search [article]

Herwig Lejsek, Björn ór Jónsson, Laurent Amsaleg, Fririk Heiar Ásmundsson
2019 arXiv   pre-print
By extending the NV-tree, a scalable disk-based high-dimensional index, we show how to implement the ACID properties of transactions which ensure both dynamicity and durability.  ...  Durability, however, has rarely been integrated within scalable and dynamic high-dimensional indexing solutions.  ...  One line of work is based on indexing data clusters such as the hierarchical k-means decomposition of the data collection: Voronoi cells are created to partition and store the high-dimensional vectors,  ... 
arXiv:1805.10942v2 fatcat:hdapj4544vhxxory4bvnxx5pzq

NV-Tree

Herwig Lejsek, Björn Þór Jónsson, Laurent Amsaleg
2011 Proceedings of the 1st ACM International Conference on Multimedia Retrieval - ICMR '11  
It addresses the specific, yet important, problem of efficiently and effectively finding the approximate k-nearest neighbors within a collection of a few billion high-dimensional data points.  ...  The NV-Tree is a very compact index, as only six bytes are kept in the index for each high-dimensional descriptor.  ...  Large Scale Approximate Indexing The literature contains several very interesting approximate high-dimensional indexing schemes.  ... 
doi:10.1145/1991996.1992050 dblp:conf/mir/LejsekJA11 fatcat:hesva7eem5emfivgtxxtyfp3cy

Adaptive Cluster Distance Bounding for High-Dimensional Indexing

Sharadh Ramaswamy, Kenneth Rose
2011 IEEE Transactions on Knowledge and Data Engineering  
We propose a new cluster-adaptive distance bound based on separating hyperplane boundaries of Voronoi clusters to complement our cluster based index.  ...  We note that indexing by "vector approximation" (VA-File), which was proposed as a technique to combat the "Curse of Dimensionality", employs scalar quantization, and hence necessarily ignores dependencies  ...  The feature vectors in a partition are now indexed by their centroid-distance, using ubiquitous one-dimensional indexes such as the B + -tree [1] .  ... 
doi:10.1109/tkde.2010.59 fatcat:2o26uoj4zvd6le2phyeujlm6x4

A Survey on Big IoT Data Indexing: Potential Solutions, Recent Advancements, and Open Issues

Zineddine Kouahla, Ala-Eddine Benrazek, Mohamed Amine Ferrag, Brahim Farou, Hamid Seridi, Muhammet Kurulay, Adeel Anjum, Alia Asheralieva
2021 Future Internet  
However, efficient retrieval and management of such information in terms of index size and search time require optimization of indexing schemes which is rather difficult to implement.  ...  The purpose of this paper is to examine and review existing indexing techniques for large-scale data.  ...  The authors also presented a comparative study of multidimensional indexing methods and a comparative study of metric access methods.  ... 
doi:10.3390/fi14010019 fatcat:xnlzg7cs2fb3lgng65ha5ucf5m

Adaptive nearest neighbor search for relevance feedback in large image databases

P. Wu, B. S. Manjunath
2001 Proceedings of the ninth ACM international conference on Multimedia - MULTIMEDIA '01  
The main contribution of this paper is a novel algorithm for adaptive nearest neighbor computations for high dimensional feature vectors and when the number of items in the database is large.  ...  Such nearest neighbor computations are expensive given that typical image features, such as color and texture, are represented in high dimensional spaces.  ...  An approximation based index scheme to support relevance feedback Typical audio-visual descriptors are high dimensional vectors [1, 11] . Their dimensionality range from few tens to a few hundreds.  ... 
doi:10.1145/500141.500157 fatcat:gp6mtdhofzfoxkcff3c2eu7eiy

Adaptive Cluster-Distance Bounding for Nearest Neighbor Search in Image Databases

Sharadh Ramaswamy, Kenneth Rose
2007 2007 IEEE International Conference on Image Processing  
We consider approaches for exact similarity search in a high dimensional space of correlated features representing image datasets, based on principles of clustering and vector quantization.  ...  We develop an adaptive cluster distance bound based on separating hyperplanes, that complements our index in selectively retrieving clusters that contain data entries closest to the query.  ...  Navneet Panda, Vinay Melkote and Ankur Saxena for their valuable suggestions. The authors are extremely grateful to Balakrishnan Selvam for his help in performing the experiments.  ... 
doi:10.1109/icip.2007.4379601 dblp:conf/icip/RamaswamyR07 fatcat:ne7teonnlzgjna6c5vh7t4msyi
« Previous Showing results 1 — 15 out of 35,192 results