Filters








1,902 Hits in 4.7 sec

Multiscale Quantization for Fast Similarity Search

Xiang Wu, Ruiqi Guo, Ananda Theertha Suresh, Sanjiv Kumar, Daniel N. Holtmann-Rice, David Simcha, Felix X. Yu
2017 Neural Information Processing Systems  
We propose a multiscale quantization approach for fast similarity search on large, high-dimensional datasets.  ...  This is a common scenario for realworld datasets, especially when doing product quantization of residuals obtained from coarse vector quantization.  ...  We thank Jeffrey Pennington and Chen Wang for their helpful comments and discussions.  ... 
dblp:conf/nips/WuGSKHSY17 fatcat:syjberesxzdcffkm4h4jh6diem

Fast image search on a VQ compressed image database

Mehmet YAKUT
2016 Turkish Journal of Electrical Engineering and Computer Sciences  
A fast and efficient image search method is developed for a compressed image database using vector quantization (VQ).  ...  The same is valid for the object search in the image database.  ...  Multiscale image representation and progressive image transmission is a very important subject when it is required for featured images to scan or a fast image search in a large image database and also  ... 
doi:10.3906/elk-1502-164 fatcat:ju2fzpdivnepjhaix5hranuk5u

Low bit-rate efficient compression for seismic data

A.Z. Averbuch, F. Meyer, J.-O. Stromberg, R. Coifman, A. Vassiliou
2001 IEEE Transactions on Image Processing  
Up to now some of the algorithms used for seismic data compression were based on some form of wavelet or local cosine transform, while using a uniform or quasiuniform quantization scheme and they finally  ...  Some of these algorithms are also suitable for multimedia type compression.  ...  Compression algorithms which work well for other types of data, such as algorithms similar to EZW and SPIHT compression schemes, perform poorly on seismic data.  ... 
doi:10.1109/83.974565 pmid:18255520 fatcat:zp6jlcmilbgofht3rb3ubrqxay

A Technique for Producing Scalable Color-Quantized Images With Error Diffusion

Y.-H. Fung, Y.-H. Chan
2006 IEEE Transactions on Image Processing  
In this paper, a color quantization algorithm for generating scalable colorindexed images is proposed based on a multiscale error diffusion framework.  ...  Index Terms-Color index, color quantization, directional hysteresis, error diffusion, halftoning, multiscale processing, scalable media.  ...  Searching the Pixel for Color Quantization The location of a pixel to be color quantized is determined via maximum energy guidance with energy pyramid E.  ... 
doi:10.1109/tip.2006.877480 pmid:17022284 fatcat:preff2teqzgmnblwxmpyyqfdbe

Low-bit-rate efficient compression for seismic data

Amir Z. Averbuch, Francois G. Meyer, Jan-Olov Stroemberg, Ronald R. Coifman, Anthony A. Vassiliou, Andrew F. Laine, Michael A. Unser, Akram Aldroubi
2001 Wavelets: Applications in Signal and Image Processing IX  
Up to now some of the algorithms used for seismic data compression were based on some form of wavelet or local cosine transform, while using a uniform or quasiuniform quantization scheme and they finally  ...  Some of these algorithms are also suitable for multimedia type compression.  ...  Compression algorithms which work well for other types of data, such as algorithms similar to EZW and SPIHT compression schemes, perform poorly on seismic data.  ... 
doi:10.1117/12.449714 fatcat:wgvt2gxtqzbc5akcglge2er6cu

Dithered Color Quantization

J.M. Buhmann, D.W. Fellner, M. Held, J. Ketterer, J. Puzicha
1998 Computer graphics forum (Print)  
A highly efficient algorithm for optimization based on a multiscale method is developed for the dithered color quantization cost function.  ...  Quantization and dithering are generally performed sequentially 1, 2, 12, 13 . It is a key observation that quantization  ...  dithered quantization is then defined as a search for a parameter set (M,Y) which minimizes (5) .  ... 
doi:10.1111/1467-8659.00269 fatcat:3eofmyrwcbf7nf5ofmihqoolfa

Reducing the Complexity of Multiscale Error Diffusion

Ka-Chun Lui, Yuk-Hee Chan
2006 TENCON 2006 - 2006 IEEE Region 10 Conference  
However, the complexity of this frame-oriented process is much higher and makes it not suitable for real-time applications. In this paper, a fast MED algorithm is proposed.  ...  Multiscale error diffusion (MED) is superior to conventional error diffusion algorithms as it can eliminate directional hysteresis completely.  ...  This paper presents an alternative to realize multiscale error diffusion.  ... 
doi:10.1109/tencon.2006.344062 fatcat:4pjzj6yqujc6feiv3fxfch54y4

Local Orthogonal Decomposition for Maximum Inner Product Search [article]

Xiang Wu, Ruiqi Guo, Sanjiv Kumar, David Simcha
2019 arXiv   pre-print
In such a framework, vector quantization is used for coarse partitioning while product quantization is used for quantizing residuals.  ...  More specifically, we decompose a residual vector locally into two orthogonal components and perform uniform quantization and multiscale quantization to each component respectively.  ...  {v i } [m] , uniform quantization {φ i U Q (·)} [m] and multiscale quantization {φ i M SQ (·)} [m] Table 2 : 2 Datasets used for MIPS experiments.  ... 
arXiv:1903.10391v1 fatcat:66mh46pzj5fjfjuad45y6nh2ca

A low-complexity multiscale error diffusion algorithm for digital halftoning

King-Hong Chung, Yik-Hing Fung, Ka-Chun Lui, Yuk-Hee Chan
2005 2005 International Symposium on Intelligent Signal Processing and Communication Systems  
However, extremely large computation effort is required for its implementation.  ...  Multiscale error diffusion (MED) digital halftoning technique outperforms classical conventional error diffusion techniques as it can produce a directionalhysteresis-free bi-level image.  ...  However, heavy computation effort is required to search the locations for error diffusion especially when the image size is large.  ... 
doi:10.1109/ispacs.2005.1595499 fatcat:pgqjncvkqzgc7fggzenrvielri

Multiscale modeling and estimation of motion fields for video coding

P. Moulin, R. Krishnamurthy, J.W. Woods
1997 IEEE Transactions on Image Processing  
We formulate the estimation and quantization problems jointly as a discrete optimization problem and solve it using a fast multiscale relaxation algorithm.  ...  It determines not only the coe cients of the model, but also the quantization method.  ...  The authors are grateful to Alex Loui, John Woods, and Jelena Kova cevi c for their helpful interaction and suggestions at various stages of this research.  ... 
doi:10.1109/83.650115 pmid:18285232 fatcat:bz47zzejhjcj7nzv6rt66b4mka

A compressive sensing approach to object-based surveillance video coding

Divya Venkatraman, Anamitra Makur
2009 2009 IEEE International Conference on Acoustics, Speech and Signal Processing  
The techniques are studied and analyzed by varying the different trade-off parameters such as the measurement index, quantization levels etc.  ...  Finally we recommend an optimal scheme for a range of bitrates. Experimental results with comparative bitrates-vs-PSNR graphs for the different techniques are presented 1  ...  Multiscale wavelet CS: For better representation of the wavelet coefficients, wavelet bands at different levels are no longer merged but treated separately in a fashion similar to [11] .  ... 
doi:10.1109/icassp.2009.4960383 dblp:conf/icassp/VenkatramanM09 fatcat:mpbzgdnkgzevhcewstiq26c354

AUDIO SIGNAL REPRESENTATIONS FOR FACTORIZATION IN THE SPARSE DOMAIN

Manuel Moussallam, Laurent Daudet, Gael Richard
2011 2011 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)  
The main feature of these representations is that they are both sparse and approximately shift-invariant, which allows similarity search in a sparse domain.  ...  In this paper, a new class of audio representations is introduced, together with a corresponding fast decomposition algorithm.  ...  For MP and LOMP decompositions, atom coefficients are encoded using a simple 7-bit uniform scalar quantizer and plain entropy coding of the quantized coefficients, atom indexes have a fixed cost of log  ... 
doi:10.1109/icassp.2011.5946453 dblp:conf/icassp/MoussallamDR11 fatcat:htgzt2obh5fm3fnltzsts4xryi

International Conference on Image Processing

1996 Proceedings of 3rd IEEE International Conference on Image Processing  
design of vector-quantized multiresolution codecs On entropy coded and entropy constrained lattice vector quantization Fast nearest neighbor search for ECVQ and other modified distorsion measures ~ Improving  ...  binary position coding Fast Wavelet transform for color image compression Direct processing of EZW compressed image data Enhanced zerotree wavelet transform image coding exploiting similarities  ... 
doi:10.1109/icip.1996.560353 fatcat:le3ysy6wxrfr7nq56ueropy7tu

Projective Clustering Product Quantization [article]

Aditya Krishnan, Edo Liberty
2021 arXiv   pre-print
This paper suggests the use of projective clustering based product quantization for improving nearest neighbor and max-inner-product vector search (MIPS) algorithms.  ...  We provide anisotropic and quantized variants of projective clustering which outperform previous clustering methods used for this problem such as ScaNN.  ...  Multiscale quantization for fast similarity search. Advances in Neural Information Processing Systems, 30:5745–5755, 2017. [45] Ting Zhang, Chao Du, and Jingdong Wang.  ... 
arXiv:2112.02179v1 fatcat:yxxazxoxlne7hkn5bsnvw3iwhu

Texture Regimes for Entropy-Based Multiscale Image Analysis [chapter]

Sylvain Boltz, Frank Nielsen, Stefano Soatto
2010 Lecture Notes in Computer Science  
We present an approach to multiscale image analysis.  ...  At each point, multiple small and large regions co-exist at multiple scales, as image structures are pooled by the scaling and quantization process to form "textures" and then transitions between textures  ...  What is critical is the composition of scaling with quantization, which makes for a semi-group.  ... 
doi:10.1007/978-3-642-15558-1_50 fatcat:u5k2kyzganecflncatb6ylunku
« Previous Showing results 1 — 15 out of 1,902 results