1,332 Hits in 5.2 sec

Som-Based K-Nearest Neighbors Search in Large Image Databases [chapter]

Zaher Aghbari, Kun-Seok Oh, Yaokai Feng, Akifumi Makinouchi
2002 Visual and Multimedia Information Management  
We address the problem of K-Nearest Neighbors (KNN) search in large image databases. Our approach is to cluster the database of n points (i.e. images) using a self-organizing map algorithm.  ...  Unlike most indexes of KNN algorithms that require storage space exponential in dimensions, the arrayindex requires a storage space that is linear in the number of generated clusters.  ...  K-NEAREST NEIGHBORS SEARCH Given a query image q, we want to retrieve the exact KNN images.  ... 
doi:10.1007/978-0-387-35592-4_5 fatcat:6vrzmnxdw5ajxdvdmf6frqnyvy

Similarity Retrieval Based on SOM-Based R*-Tree [chapter]

K. H. Choi, M. H. Shin, S. H. Bae, C. H. Kwon, I. H. Ra
2004 Lecture Notes in Computer Science  
Feature-based similarity retrieval has become an important research issue in multimedia database systems. The features of multimedia data are usually high-dimensional data.  ...  In this paper, we propose a SOM-based R*-tree(SBR-Tree) as a new indexing method for high-dimensional feature vectors.  ...  The data set size was from 1,000 to 40,000 images. The search method used was the κ-Nearest Neighbor(NN)[15] method, which searches for κ (κ>1) objects nearest to the given query.  ... 
doi:10.1007/978-3-540-24688-6_33 fatcat:ds3oizmurnhlbg46zi4dpzdwb4

SOM-based R*-tree for similarity retrieval

Kun-Seok Oh, Yaokai Feng, K. Kaneko, A. Makinouchi, Sang-Hyun Bae
2001 Proceedings Seventh International Conference on Database Systems for Advanced Applications DASFAA 2001 DASFAA-01  
Feature-based similarity retrieval has become an iniportant research issue in multimedia database systems.  ...  The SOM-based R*-tree combines SOM and R*tree to achieve search performance more scalable to high dimensionalities.  ...  This work was supported in part by a Grant-in-Aid for Scientific Research (10308012) from the Ministry of Education, Science, Sports and Culture of Japan.  ... 
doi:10.1109/dasfaa.2001.916377 dblp:conf/dasfaa/OhFKMB01 fatcat:grehhoy4kzeetkwebc35skj7ei

Deep Supervised Quantization by Self-Organizing Map

Min Wang, Wengang Zhou, Qi Tian, Junfu Pu, Houqiang Li
2017 Proceedings of the 2017 ACM on Multimedia Conference - MM '17  
Approximate Nearest Neighbour (ANN) search is an important research topic in multimedia and computer vision fields.  ...  In this paper, we propose a new deep supervised quantization method by Self-Organizing Map (SOM) to address this problem.  ...  Given a query, ANN search aims to find its nearest neighbours from a large-scale dataset in a sublinear, or even constant time complexity.  ... 
doi:10.1145/3123266.3123415 dblp:conf/mm/WangZTPL17 fatcat:4kjtvt7e7bb5ravxckg6jpphgq

Content-Based Image Retrieval Using Self-Organizing Maps [chapter]

Jorma Laaksonen, Markus Koskela, Erkki Oja
1999 Lecture Notes in Computer Science  
In many settings, the existing and widely adopted methods for text-based indexing and information retrieval are inadequate for these new purposes.  ...  Digital image libraries are becoming more common and widely used as visual information is produced at a rapidly growing rate.  ...  They applied the SOM as a filter of unlikely relevant images based on color and texture features. Within the images mapped to the BMU, a search for k nearest neighbors is then performed.  ... 
doi:10.1007/3-540-48762-x_67 fatcat:6pqbfjjhibbudlx5fquttmzete

Motion Recurring Pattern Analysis: A Lossless Representation for Motion Capture Databases

Pengjie Wang, Jiang Wang, Xiaoming Wei, Jiana Meng, Jing Xun
2020 IEEE Access  
First, we concatenate all the motions into a long sequence in the motion database, and we discover similar posture paths by building a matching trellis structure based on the randomized k-d tree.  ...  In this paper, we propose the motion recurring pattern analysis (MRPA) method for the lossless representation of a motion database at the segment level instead of the motion degree of freedom (DOF) level  ...  As defined in [47] , there are two forms of the nearest neighbor search based on the randomized k-d tree structure: the k-nearest neighbor search and the radius nearest neighbor search.  ... 
doi:10.1109/access.2020.2989430 fatcat:cp2hhce4nzdf3p2piqroych22e

International Research Journal of Engineering Science, Technology and Innovation

2014 International Research Journal of Engineering Science Technology and Innovation  
Here we also studied an image indexing method based on Self-Organizing Maps (SOM). The SOM was interpreted as a combination of clustering and dimensionality reduction.  ...  It contains the problem of finding images relevant to the users' information needs from image databases, based principally on low-level visual features for which automatic extraction methods are available  ...  As all images in the database have been previously mapped in their best-matching SOM units at the time the SOMs were trained, and easy to locate the positive and negative images on each SOM in use.  ... 
doi:10.14303/irjesti.2014.054 fatcat:vdhegqs2xnaini56vjemik4fha

Active browsing using similarity pyramids

Jau-Yuen Chen, Charles A. Bouman, John C. Dalton, Minerva M. Yeung, Boon-Lock Yeo, Charles A. Bouman
1998 Storage and Retrieval for Image and Video Databases VII  
In this paper, we describe a new approach to managing large image databases which we call active browsing.  ...  At coarse levels, the similarity pyramid allows users to view the database as large clusters of similar images. Alternatively, users can "zoom into" finer levels to view individual images.  ...  INTRODUCTION Many approaches for content based management of image databases have focused on query-by-example methods 1 in which an example image is presented and the database is searched for images with  ... 
doi:10.1117/12.333834 dblp:conf/spieSR/ChenBD99 fatcat:4lmskbsfpnfbfgbv7aapvmus5q

Content-based image retrieval using growing hierarchical self-organizing quadtree map

Sitao Wu, M.K.M. Rahman, Tommy W.S. Chow
2005 Pattern Recognition  
We also propose feature matrices, image distance and relevance feedback for region-based images in the GHSOQM-based CBIR system.  ...  In this paper, a growing hierarchical self-organizing quadtree map (GHSOQM) is proposed and used for a content-based image retrieval (CBIR) system.  ...  The work described in this paper was fully supported by a grant from the Research Grant Council of Hong Kong SAR Region [Project No. 7001599-570].  ... 
doi:10.1016/j.patcog.2004.10.005 fatcat:zor77lku75ggzh4xdnezjf4kyi

Exploring objects for recognition in the real word

Gert Kootstra, Jelmer Ypma, Bart de Boer
2007 2007 IEEE International Conference on Robotics and Biomimetics (ROBIO)  
Furthermore, we propose a model to reduce the number of SIFT keypoints stored in the object database.  ...  We discuss the use of a growing-when-required (GWR) network, which is based on the Kohonen Self Organizing Feature Map, for efficient clustering of the keypoints.  ...  The activation of a model is based on the set of observations, M , that support the modelare respectively the object ID and pose of the nearest neighbor, n k , of i p in the keypoint database.  ... 
doi:10.1109/robio.2007.4522200 dblp:conf/robio/KootstraYB07 fatcat:kcuxs3zch5fehmpdtuxmgnj57q

A Review on Clustering Technique

Vivek Kumar
2015 International Journal on Recent and Innovation Trends in Computing and Communication  
Large data set have many hidden pattern which have very crucial information, Clustering is such technique which find the hidden pattern from the large data.  ...  Example of Competitive learning, SOM and ART are famous for clustering. SOM have the limitation of dimension, ART is good but computation cost is very high.  ...  Data mining derives its name from the similarities between searching for valuable information in a large database.  ... 
doi:10.17762/ijritcc2321-8169.1503136 fatcat:rdt7mpmyvnbm3ph6o7v72wmpdy

An Indexing and Retrieval Mechanism for Complex Similarity Queries in Image Databases

Guang-Ho Cha, Chin-Wan Chung
1999 Journal of Visual Communication and Image Representation  
A content-based image retrieval mechanism to support complex similarity queries is presented.  ...  Experiments have been carried out on large image collections to demonstrate the effectiveness of the proposed retrieval mechanism.  ...  X i will contain the objects // returned from the k-nearest neighbor search. } 2.  ... 
doi:10.1006/jvci.1999.0410 fatcat:kg4p74nuxvg7diezg5jynizns4

Recognizing Partially Occluded, Expression Variant Faces From Single Training Image per Person With SOM and Soft$k$-NN Ensemble

X. Tan, S. Chen, Z.-H. Zhou, F. Zhang
2005 IEEE Transactions on Neural Networks  
A soft k nearest neighbor (soft k-NN) ensemble method, which can effectively exploit the outputs of the SOM topological space, is also proposed to identify the unlabelled subjects.  ...  Most classical template-based frontal face recognition techniques assume that multiple images per person are available for training, while in many real-world applications only one training image per person  ...  Martinez for providing us the AR database. Portions of the research in this paper use the FERET database of facial images collected under the FERET program.  ... 
doi:10.1109/tnn.2005.849817 pmid:16121729 fatcat:rjdw27d6xbcvnmzr5xvdtew6qi

Supervised Image Classification by SOM Activity Map Comparison

G. Lefebvre, C. Laurent, J. Ros, C. Garcia
2006 18th International Conference on Pattern Recognition (ICPR'06)  
In this paper, we propose to use self-organizing feature maps (SOM) to measure image similarity. To reach this goal, we feed local signatures associated to salient patches into the neural network.  ...  This kind of problem is recurrent in many applications such as object recognition, image classification, etc.  ...  , we compute the k-nearest learning images based on the previous criterion (5) and a k-nn classification is performed.  ... 
doi:10.1109/icpr.2006.1094 dblp:conf/icpr/LefebvreLRG06 fatcat:yxrx3tmkrbavbd5763dqp7y7em

Defect Image Classification and Retrieval with MPEG-7 Descriptors [chapter]

Jussi Pakkanen, Antti Ilvesmäki, Jukka Iivarinen
2003 Lecture Notes in Computer Science  
In this paper the visual content descriptors defined by the MPEG-7 standard are applied to defect image classification and retrieval. A pre-classified defect image database is used in evaluation.  ...  The experiments are done with a KNN classifier and with a PicSOM content-based image retrieval system.  ...  KNN Cross-Validation The performance of the descriptors was first evaluated by performing a leaveone-out k-nearest neighbor (KNN) cross-validatory classification with k=5.  ... 
doi:10.1007/3-540-45103-x_48 fatcat:mj4siab7uvg3tc6egurybphvey
« Previous Showing results 1 — 15 out of 1,332 results