Filters








4,807 Hits in 7.9 sec

Fast approximate spectral clustering

Donghui Yan, Ling Huang, Michael I. Jordan
2009 Proceedings of the 15th ACM SIGKDD international conference on Knowledge discovery and data mining - KDD '09  
Spectral clustering refers to a flexible class of clustering procedures that can produce high-quality clusterings on small data sets but which has limited applicability to large-scale problems due to its  ...  We extend the range of spectral clustering by developing a general framework for fast approximate spectral clustering in which a distortion-minimizing local transformation is first applied to the data.  ...  Fast spectral clustering with k-means Vector quantization is the problem of choosing a set of representative points that best represent a data set in the sense of minimizing a distortion measure [13]  ... 
doi:10.1145/1557019.1557118 dblp:conf/kdd/YanHJ09 fatcat:ltxmgx2vrfgnzlsg527w7decda

Quantization of Product using Collaborative Filtering Based on Cluster

E. Sankar, L. Karthik, Kuppa Venkatasriram Sastry
2022 International Journal for Research in Applied Science and Engineering Technology  
Locality sensitive hashing and index-based methods usually store both index data and item feature vectors in main memory, so they handle a limited number of items.  ...  Hashing-based recommendation methods enjoy low memory cost and fast retrieval of items, but suffer from large accuracy degradation.  ...  Most importantly, unlike hashing-based methods representing each data instance by a hash code, which depends on a set of hash functions, quantization based methods represent each data instance by an index  ... 
doi:10.22214/ijraset.2022.40753 fatcat:pnsv22xv3vbnzndibgfuakeb4y

Compression of Spectral Images [chapter]

Arto Kaarna
2007 Vision Systems: Segmentation and Pattern Recognition  
In vector quantization each vector is represented by the centroid of a cluster it belongs to.  ...  Vector quantization Clustering is an unsupervised method to classify patterns in an image.  ...  The later chapters are devoted to pattern recognition and covers diverse topics ranging from biological image analysis, remote sensing, text recognition, advanced filter design for data analysis, etc.  ... 
doi:10.5772/4964 fatcat:dulzntdrqfcrvelgboxeohl53i

Multi-level Low-rank Approximation-based Spectral Clustering for image segmentation

Lijun Wang, Ming Dong
2012 Pattern Recognition Letters  
In this paper, we propose an efficient spectral method, Multi-level Low-rank Approximation-based Spectral Clustering (MLASC), to segment high resolution images.  ...  Spectral clustering is a well-known graph-theoretic approach of finding natural groupings in a given dataset, and has been broadly used in image segmentation.  ...  , INyström with k-means as a preprocessor , and KASP (Yan et al., 2009 ), a quantization-based fast spectral clustering method.  ... 
doi:10.1016/j.patrec.2012.07.024 fatcat:f4caitkrrreedjokuzpipslunu

Learning Better Encoding for Approximate Nearest Neighbor Search with Dictionary Annealing [article]

Shicong Liu, Hongtao Lu
2015 arXiv   pre-print
We introduce a novel dictionary optimization method for high-dimensional vector quantization employed in approximate nearest neighbor (ANN) search.  ...  Vector quantization methods first seek a series of dictionaries, then approximate each vector by a sum of elements selected from these dictionaries.  ...  For a large-scale dataset, performing k-means on all data could be prohibitive, and the size of datasets may grow with time.  ... 
arXiv:1507.01442v1 fatcat:64deguaupbh2lci4ppvvdjnqmu

Approximate spectral clustering using both reference vectors and topology of the network generated by growing neural gas [article]

Kazuhisa Fujita
2021 arXiv   pre-print
Spectral clustering (SC) is one of the most popular clustering methods and often outperforms traditional clustering methods.  ...  ASC with GNG uses not only reference vectors for vector quantization but also the topology of the network for extraction of the topological relationship between data points in a dataset.  ...  Spectral clustering (SC) is a popular modern clustering method based on eigendecomposition of a Laplacian matrix calculated from a similarity matrix of a dataset (Taşdemir et al., 2015) .  ... 
arXiv:2009.07101v4 fatcat:fdilfoabefc7pdnus6w7tvslzy

LOH and behold: Web-scale visual search, recommendation and clustering using Locally Optimized Hashing [article]

Yannis Kalantidis, Lyndon Kennedy, Huy Nguyen, Clayton Mellina, David A. Shamma
2016 arXiv   pre-print
We propose a novel hashing-based matching scheme, called Locally Optimized Hashing (LOH), based on a state-of-the-art quantization algorithm that can be used for efficient, large-scale search, recommendation  ...  In this paper we experiment on datasets of up to 100 million images, but in practice our system can scale to larger collections and can be used for other types of data that have a vector representation  ...  Related Work Large scale nearest neighbor search was traditionally based on hashing methods [6, 26] because they offer low memory footprints for index codes and fast search in Hamming space [25] .  ... 
arXiv:1604.06480v2 fatcat:odskuod7cbgndjezmuz77s64fe

LOH and Behold: Web-Scale Visual Search, Recommendation and Clustering Using Locally Optimized Hashing [chapter]

Yannis Kalantidis, Lyndon Kennedy, Huy Nguyen, Clayton Mellina, David A. Shamma
2016 Lecture Notes in Computer Science  
We propose a novel hashing-based matching scheme, called Locally Optimized Hashing (LOH), based on a state-of-the-art quantization algorithm that can be used for efficient, large-scale search, recommendation  ...  In this paper we experiment on datasets of up to 100 million images, but in practice our system can scale to larger collections and can be used for other types of data that have a vector representation  ...  Related Work Large scale nearest neighbor search was traditionally based on hashing methods [6, 26] because they offer low memory footprints for index codes and fast search in Hamming space [25] .  ... 
doi:10.1007/978-3-319-46604-0_49 fatcat:ee4xpo3aprh5pf6q374pcztdvy

Scalable Image Retrieval by Sparse Product Quantization [article]

Qingqun Ning, Jianke Zhu, Zhiyuan Zhong, Steven C.H. Hoi, Chun Chen
2016 arXiv   pre-print
We obtain state-of-the-art results for ANN search on four public image datasets and the promising results of content-based image retrieval further validate the efficacy of our proposed method.  ...  Fast Approximate Nearest Neighbor (ANN) search technique for high-dimensional feature indexing and retrieval is the crux of large-scale image retrieval.  ...  A. Vector Quantization Vector Quantization (VQ) [13] is a classical technique for data compression.  ... 
arXiv:1603.04614v1 fatcat:mskz4ksvfrcjfkdjiujtfighki

Scalable Image Retrieval by Sparse Product Quantization

Qingqun Ning, Jianke Zhu, Zhiyuan Zhong, Steven C.H. Hoi, Chun Chen
2017 IEEE transactions on multimedia  
We obtain state-of-the-art results for ANN search on four public image datasets and the promising results of content-based image retrieval further validate the efficacy of our proposed method.  ...  Fast Approximate Nearest Neighbor (ANN) search technique for high-dimensional feature indexing and retrieval is the crux of large-scale image retrieval.  ...  A. Vector Quantization Vector Quantization (VQ) [13] is a classical technique for data compression.  ... 
doi:10.1109/tmm.2016.2625260 fatcat:d4kllkspz5fvzdmruw6hsnzqra

Approximate spectral clustering using both reference vectors and topology of the network generated by growing neural gas

Kazuhisa Fujita
2021 PeerJ Computer Science  
Spectral clustering (SC) is one of the most popular clustering methods and often outperforms traditional clustering methods.  ...  ASC with GNG uses not only reference vectors for vector quantization but also the topology of the network for extraction of the topological relationship between data points in a dataset.  ...  Spectral clustering (SC) is a popular modern clustering method based on eigendecomposition of a Laplacian matrix calculated from a similarity matrix of a dataset (Taşdemir, Yalçin & Yildirim, 2015) .  ... 
doi:10.7717/peerj-cs.679 pmid:34497872 pmcid:PMC8384042 fatcat:57bnwjpkt5erliivxir5stkqvq

Extremely Low Bit-Rate Nearest Neighbor Search Using a Set Compression Tree

Relja Arandjelovic, Andrew Zisserman
2014 IEEE Transactions on Pattern Analysis and Machine Intelligence  
The goal of this work is a data structure to support approximate nearest neighbor search on very large scale sets of vector descriptors.  ...  The method is compared on standard benchmarks (SIFT1M and 80 Million Tiny Images) to a number of state of the art approaches, including Product Quantization, Locality Sensitive Hashing, Spectral Hashing  ...  We are grateful for financial support from ERC grant VisRec no. 228180.  ... 
doi:10.1109/tpami.2014.2339821 pmid:26353147 fatcat:qnt6vtbt35dddohgoryora7h6e

Web-Scale Image Clustering Revisited

Yannis Avrithis, Yannis Kalantidis, Evangelos Anagnostopoulos, Ioannis Z. Emiris
2015 2015 IEEE International Conference on Computer Vision (ICCV)  
Principled clustering methods, especially kernelized and spectral ones, have higher complexity and are difficult to scale above millions.  ...  Large scale duplicate detection, clustering and mining of documents or images has been conventionally treated with seed detection via hashing, followed by seed growing heuristics using fast search.  ...  We thank Clayton Mellina and the Flickr Vision Team for their help with CNN features and the distributed k-means experiment.  ... 
doi:10.1109/iccv.2015.176 dblp:conf/iccv/AvrithisKAE15 fatcat:xbsfi2ktsnhipb5tl5lwz4g26e

Neural network-based clustering for agriculture management

Kadim Taşdemir, Csaba Wirnhardt
2012 EURASIP Journal on Advances in Signal Processing  
We propose such a method using self-organizing maps (SOM) based spectral clustering, for agriculture management.  ...  By combining the powerful aspects of the SOM (adaptive vector quantization in a topology preserving manner) and of the spectral clustering (a manifold learning based on eigendecomposition of pairwise similarities  ...  providing us LPIS vector data and orthophotos for this study.  ... 
doi:10.1186/1687-6180-2012-200 fatcat:uv6hnjj2e5b37dnjdl3r27nhjy

Generalized residual vector quantization for large scale data [article]

Shicong Liu, Junru Shao, Hongtao Lu
2016 arXiv   pre-print
Vector quantization is an essential tool for tasks involving large scale data, for example, large scale similarity search, which is crucial for content-based information retrieval and analysis.  ...  First, we provide a detailed review on a relevant vector quantization method named residual vector quantization (RVQ).  ...  We train all methods on the training set and encode the base dataset. We train online version of GRVQ with all the data. The training time on GIST1M of all methods is presented on Fig.2(e) .  ... 
arXiv:1609.05345v1 fatcat:yiu4dx27bnewrd6xe5z2eljuw4
« Previous Showing results 1 — 15 out of 4,807 results