2,117 Hits in 7.2 sec

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

Romain Tavenard, Herve Jegou, Laurent Amsaleg
2011 2011 9th International Workshop on Content-Based Multimedia Indexing (CBMI)  
Experiments with a large scale collection of image descriptors show that our algorithm significantly reduces the variance of response times without severely impacting the search quality.  ...  Clusters are often produced by the well-known k-means approach since it has several desirable properties. On the downside, it tends to produce clusters having quite different cardinalities.  ...  Acknowledgements This work was partly realized as part of the Quaero Project, funded by OSEO, French State agency for innovation.  ... 
doi:10.1109/cbmi.2011.5972514 dblp:conf/cbmi/TavenardJA11 fatcat:kmslctexprbznlvznjtwmuyeeu

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

Romain Tavenard , Hervé Jégou
2010 arXiv   pre-print
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.  ...  Clusters are often produced by the well-known $k$-means approach since it has several desirable properties. On the downside, it tends to produce clusters having quite different cardinalities.  ...  Section 4 evaluates the impact of balancing on the response time of queries when using large collections of descriptors computed over 1 million images from Flickr.  ... 
arXiv:1009.4739v1 fatcat:hdod6pwlgbbwnpwhfvhiel45vm

A Probabilistic Analysis of Sparse Coded Feature Pooling and Its Application for Image Retrieval

Yunchao Zhang, Jing Chen, Xiujie Huang, Yongtian Wang, Rongrong Ji
2015 PLoS ONE  
Further we apply sparse coding method to aggregate multiple types of features for large-scale image retrieval.  ...  In this paper, we first analyze the effects of different sampling strategies for image retrieval, then we discuss feature pooling strategies on image retrieval performance with a probabilistic explanation  ...  problem of large-scale image retrieval.  ... 
doi:10.1371/journal.pone.0131721 pmid:26132080 pmcid:PMC4489107 fatcat:7gezpztqsrbhfb5wjfk6wfad7m

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  
The created feature space is then used, on one hand to approximate regions, and on the other hand to provide an effective kernel distances for both filtering process and similarity measurement.  ...  However, these methods do not generally support efficiently similarity search when dealing with heterogeneous data vectors.  ...  Acknowledgement This work was partly supported by the STIC France-Morocco program (Sciences et Technologies de l'Information et de la Communication).  ... 
doi:10.1016/j.image.2009.09.001 fatcat:v5ce4yqlw5en5gc5sbwk763hru

The power of comparative reasoning

Jay Yagnik, Dennis Strelow, David A. Ross, Ruei-sung Lin
2011 2011 International Conference on Computer Vision  
These machine-learning-free methods when applied to the task of fast similarity search outperform state-of-theart machine learning methods with complex optimization setups.  ...  For solving classification problems, the embeddings provide a nonlinear transformation resulting in sparse binary codes that are well-suited for a large class of machine learning algorithms.  ...  Similarity Search in High Dimensions Image retrieval on LabelMe We begin by demonstrating the performance of WTA codes on the exact experiment described in [20, 19] .  ... 
doi:10.1109/iccv.2011.6126527 dblp:conf/iccv/YagnikSRL11 fatcat:p6sstbbi7bh53npwjj6wvkqzr4

Selecting Local Region Descriptors with a Genetic Algorithm for Real-World Place Recognition [chapter]

Leonardo Trujillo, Gustavo Olague, Francisco Fernández de Vega, Evelyne Lutton
2008 Lecture Notes in Computer Science  
The system relies on a bag of features approach using locally prominent image regions.  ...  From this set the system needs to determine which subset of descriptors captures regularities between image regions of the same location, and also discriminates between regions of different places.  ...  Research funded by the Ministerio de Educación y Ciencia (project Oplink -TIN2005-08818-C04), the LAFMI project, and the Junta de Extremadura Spain.  ... 
doi:10.1007/978-3-540-78761-7_33 fatcat:hrrz3pkevfcw5m4ukgygq43e4i

SEPIM: Secure and Efficient Private Image Matching

Zaid Abduljabbar, Hai Jin, Ayad Ibrahim, Zaid Hussien, Mohammed Hussain, Salah Abbdal, Deqing Zou
2016 Applied Sciences  
We present the development and validation of a secure scheme to measure the cosine similarity between two descriptor sets.  ...  We conducted several empirical analyses on real image collections to demonstrate the performance of our work.  ...  Conflicts of Interest: The authors declare no conflict of interest.  ... 
doi:10.3390/app6080213 fatcat:ytoi64rnkfanvcmhnf3eu534vy

Scalability of local image descriptors

Herwig Lejsek, Fridrik H. Ásmundsson, Björn Thór Jónsson, Laurent Amsaleg
2006 Proceedings of the 14th annual ACM international conference on Multimedia - MULTIMEDIA '06  
Finally, we test our descriptors on a collection of over three hundred thousand images, resulting in over 200 million local descriptors, and show that even at such a large scale the results are still of  ...  Recently, we have developed the PvS-framework, which allows efficient querying of large local descriptor collections.  ...  The descriptors are then segmented based on the value of their projection to this line into a set of new temporary sub-segments of identical cardinality.  ... 
doi:10.1145/1180639.1180760 dblp:conf/mm/LejsekAJA06a fatcat:sptzxzpb6bfubeopjzbj747pfm

LBP Channels for Pedestrian Detection

Remi Trichet, Francois Bremond
2018 2018 IEEE Winter Conference on Applications of Computer Vision (WACV)  
This type of descriptor usually selects a set of one-valued filters within the enormous set of all possible filters for improved efficiency.  ...  The main claim underpinning this paper is that the recent works on channel-based features restrict the filter space search, therefore bringing along the obsolescence of one-valued filter representation  ...  The detector speed is, of course, largely influenced by the descriptor parameters, which are the number of filters and channels.  ... 
doi:10.1109/wacv.2018.00122 dblp:conf/wacv/TrichetB18 fatcat:qdmzjx2d5ndmlnxb35nl2mdbpy

Graph Entropies in Texture Segmentation of Images [chapter]

Martin Welk
2016 Mathematical Foundations and Applications of Graph Entropy  
We study the applicability of a set of texture descriptors introduced in recent work by the author to texture-based segmentation of images.  ...  The texture descriptors under investigation result from applying graph indices from quantitative graph theory to graphs encoding the local structure of images.  ...  Here, it is important that this radius of influence is smaller than the patch size underlying the graph construction, such that the cut-off of the graphs has no significant influence on the values of the  ... 
doi:10.1002/9783527693245.ch7 fatcat:buxznt7mjjemdpoulpzqh5psy4

Thick boundaries in binary space and their influence on nearest-neighbor search

Tomasz Trzcinski, Vincent Lepetit, Pascal Fua
2012 Pattern Recognition Letters  
Binary descriptors allow faster similarity computation than real-valued ones while requiring much less storage.  ...  The problem of matching high-dimensional descriptors against large databases is pervasive in Computer Vision for applications such as image-retrieval, pose-estimation, and 3D-reconstruction.  ...  Many efficient approximate algorithms have therefore been proposed for large-scale search.  ... 
doi:10.1016/j.patrec.2012.08.006 fatcat:3znq765zt5gl7omss2khpy4zbm

Time-sensitive web image ranking and retrieval via dynamic multi-task regression

Gunhee Kim, Eric P. Xing
2013 Proceedings of the sixth ACM international conference on Web search and data mining - WSDM '13  
and learns stochastic parametric models to solve the relations between image occurrence probabilities and various temporal factors that influence them.  ...  Inspired by recently emerging interests on query dynamics in information retrieval research, our time-sensitive image retrieval algorithm can infer users' implicit search intent better and provide more  ...  Web image dynamics: This line of research aims at modeling how the contents of large-scale Web image collections change over time [12, 15, 16, 24] .  ... 
doi:10.1145/2433396.2433417 dblp:conf/wsdm/KimX13 fatcat:tobxxwsmbvfi7axjx6d3umiwne

An image-computable model of human visual shape similarity

Yaniv Morgenstern, Frieder Hartmann, Filipp Schmidt, Henning Tiedemann, Eugen Prokott, Guido Maiello, Roland W. Fleming, Ronald van den Berg
2021 PLoS Computational Biology  
Our findings show that incorporating multiple ShapeComp dimensions facilitates the prediction of human shape similarity across a small number of shapes, and also captures much of the variance in the multiple  ...  Yet, to date, no image-computable model can predict how visually similar or different shapes appear.  ...  Acknowledgments We thank Saskia Honnefeller, Jasmin Kleis, and Marcel Schepko for their help setting up the experiments and running initial pilot studies.  ... 
doi:10.1371/journal.pcbi.1008981 pmid:34061825 fatcat:ausuzxntrfajpfq5sojyydpwrm

NV-Tree: An Efficient Disk-Based Index for Approximate Search in Very Large High-Dimensional Collections

Herwig Lejsek, Friðrik Heiðar Ásmundsson, Björn Þór Jónsson, Laurent Amsaleg
2009 IEEE Transactions on Pattern Analysis and Machine Intelligence  
Finally, we compare the NV-tree to Locality Sensitive Hashing, a popular method for -distance search. We show that they return results of similar quality, but the NV-tree uses many fewer disk reads.  ...  Over the last two decades, much research effort has been spent on nearest neighbor search in high-dimensional data sets.  ...  The authors would like to thank Morgunblaðið for the use of their large picture collection, the authors of LSH and SIFT for giving them access to their implementations, and the anonymous reviewers for  ... 
doi:10.1109/tpami.2008.130 pmid:19299861 fatcat:7ofvu7ri2jfcjogpziwvldia4u

High-dimensional signature compression for large-scale image classification

Jorge Sanchez, Florent Perronnin
2011 CVPR 2011  
We address image classification on a large-scale, i.e. when a large number of images and classes are involved.  ...  Second, we tackle the problem of data compression on very large signatures (on the order of 10 5 dimensions) using two lossy compression strategies: a dimensionality reduction technique known as the hash  ...  If we use on average b bits per dimension to encode a given image signature (b might be a fractional value), then the cardinality of the codebook is 2 bE .  ... 
doi:10.1109/cvpr.2011.5995504 dblp:conf/cvpr/SanchezP11 fatcat:n6aiec3jqreh7agdpfv54curti
« Previous Showing results 1 — 15 out of 2,117 results