A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2012; you can also visit the original URL.
The file type is application/pdf.
Filters
A fractional bit encoding technique for the GMM-based block quantisation of images
2005
Digital signal processing (Print)
Preserved Fulltext
Our technique, which is based on a generalisation of positional value number systems, allows the use of level-based scalar quantisers in a fixed-rate coding framework. ...
Experimental results comparing the use of the bits-based GMM-based block quantiser with the levels-based one in image coding show a finite improvement in the PSNR performance. 2005 Elsevier Inc. ...
Since the non-uniform scalar quantisers in block quantisers are designed to quantise Gaussian sources efficiently, any PDF mismatch will invariably cause a degradation in performance. ...
doi:10.1016/j.dsp.2004.11.002
fatcat:33qasxh3y5bfne633wc264ndpu
Low-complexity GMM-based block quantisation of images using the discrete cosine transform
2005
Signal processing. Image communication
Preserved Fulltext
A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2005; you can also visit the original URL.
The file type is application/pdf.
While block transform image coding has not been very popular lately in the presence of current state-of-the-art wavelet-based coders, the Gaussian mixture model (GMM)-based block quantiser, without the ...
These advantages, together with bitrate scalability, result in a block quantiser that is considerably faster and less complex while the novelty of using a GMM to model the source probability density function ...
Since the scalar quantisers in traditional block quantisers are designed to quantise Gaussian sources efficiently, any PDF mismatch will invariably cause a degradation in performance. ...
doi:10.1016/j.image.2005.03.001
fatcat:e4nbvtbeqbbwdccz6ltd5lukwy
Scalable distributed speech recognition using Gaussian mixture model-based block quantisation
2006
Speech Communication
Preserved Fulltext
A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2006; you can also visit the original URL.
The file type is application/pdf.
In this paper, we investigate the use of block quantisers based on Gaussian mixture models (GMMs) for the coding of Mel frequency-warped cepstral coefficient (MFCC) features in distributed speech recognition ...
We have evaluated the quantisation scheme on the Aurora-2 database in a DSR framework. ...
However, the probability density functions (PDF) of real life sources are rarely Gaussian and any PDF mismatch with the quantiser will invariably cause a degradation in performance. ...
doi:10.1016/j.specom.2005.10.002
fatcat:hkcw3xygljh4pigh7ld56bjbze
Multi-frame GMM-based block quantisation of line spectral frequencies
2005
Speech Communication
Preserved Fulltext
A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2008; you can also visit the original URL.
The file type is application/pdf.
In this paper, we investigate the use of the Gaussian mixture model-based block quantiser for coding line spectral frequencies that uses multiple frames and mean squared error as the quantiser selection ...
Experiments on narrowband speech from the TIMIT database demonstrate that the multi-frame GMM-based block quantiser can achieve a spectral distortion of 1 dB at 22 bits/frame, or 21 bits/frame with some ...
In (Subramaniam and Rao, 2003) , a Gaussian mixture model (GMM) is used to parameterise the probability density function (PDF) of the source and optimised Gaussian block quantisers are designed for each ...
doi:10.1016/j.specom.2005.02.007
fatcat:k7xkglh3qfab7puaxqb2zhmk44
Physical characteristics of wireless communication channels for secret key establishment: A survey of the research
2018
Computers & security
Preserved Fulltext
A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2020; you can also visit the original URL.
The file type is application/pdf.
on nodes' harvesting abilities and the corresponding quantisation characteristics. ...
Reference [12] identifies two families of protocols, specifically based on the received signal strength (RSS) and on the channel impulse response (CIR). ...
steps based on Bose-Chaudhuri-Hocquenghem (BCH) codes and unidirectional hash functions, respectively. ...
doi:10.1016/j.cose.2018.08.001
fatcat:o3omsqut6rhyjhgkofbl2oiyee
Rapid digital architecture design of orthogonal matching pursuit
2016
2016 24th European Signal Processing Conference (EUSIPCO)
Preserved Fulltext
A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2020; you can also visit the original URL.
The file type is application/pdf.
Further, sparsity-regularised tree search algorithms will be examined as well as the multi-user data frame detection with a priori unknown user activities based on specifically designed user codes derived ...
By consequently working with data type-agnostic source codes, the RDAM brings significant simplifications to the fixed-point conversion of algorithms and the creation of complex-valued arithmetic. ...
The source code examples in Listings 4.1 and 4.2 make use of three custom-declared data types data_t, accum_t and result_t for this reason. ...
doi:10.1109/eusipco.2016.7760570
dblp:conf/eusipco/KnoopRSPP16
fatcat:bo3c4fo6njeo3hjjsnwo72jbwe
Distributed Monoview and Multiview Video Coding
2007
IEEE Signal Processing Magazine
Preserved Fulltext
A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2017; you can also visit the original URL.
The file type is application/pdf.
This set-up where one source is transmitted at full rate [e.g., R Y = H(Y )] and used as side information (SI) to decode the other one (implying R X = H(X|Y ) or reciprocally) corresponds to one of the ...
The proof of the SW theorem is based on random binning [4], which is nonconstructive, i.e., it does not reveal how practical code design should be done. ...
Practical code constructions based on the WZ theorem thus naturally rely on a quantizer (source code) followed by an SW coder (channel code). ...
doi:10.1109/msp.2007.904808
fatcat:xliten6xofdzxbnx5ockx4uipq
A Survey on Deep-Learning based Techniques for Modeling and Estimation of MassiveMIMO Channels
[article]
2020
arXiv
pre-print
Preserved Fulltext
A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2020; you can also visit the original URL.
The file type is application/pdf.
Other Versions
axis for a Stackelberg game based scenario. ...
In addition, we consequently propose 7 open trends e.g. in the context of the lack of Q-learning in MaMIMO detection - for which we talk about a possible solution to the saddle-point in the 2-D pilot-data ...
According to the universal approximation theorem, this function is estimated. ...
arXiv:1910.03390v2
fatcat:wtvvfve5zfhjjhicbcjzfoy3du
Optimal forgetting: Semantic compression of episodic memories
2020
PLoS Computational Biology
Preserved Fulltext
A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2020; you can also visit the original URL.
The file type is application/pdf.
We use three datasets, chess games, natural text, and hand-drawn sketches, to demonstrate the effects of semantic compression on memory performance. ...
Our model accounts for memory distortions related to domain expertise, gist-based distortions, contextual effects, and delayed recall. ...
Acknowledgments The authors would like to thank Mihály Bányai and Csenge Fráter for discussions and comments on the manuscript.
Author Contributions Conceptualization: David G. ...
doi:10.1371/journal.pcbi.1008367
pmid:33057380
fatcat:ub6wvtyi2raotaxfrtvwloaswa
Optimal forgetting: Semantic compression of episodic memories
[article]
2020
bioRxiv
pre-print
Preserved Fulltext
A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2020; you can also visit the original URL.
The file type is application/pdf.
We use three datasets, chess games, natural text, and hand-drawn sketches, to demonstrate the effects of semantic compression on memory performance. ...
Our model accounts for memory distortions related to domain expertise, gist-based distortions, contextual effects, and delayed recall. ...
Acknowledgements The authors would like to thank Mihály Bányai and Csenge Fráter for discussions and comments on the manuscript and to the anonymous reviewers for their constructive criticism. ...
doi:10.1101/2020.05.06.080838
fatcat:po4criizfvbczk2oj4hqi3sija
Time-domain coding of (near) toll quality speech at rates below 16 KB/S
1985
Signal Processing
Preserved Fulltext
A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2019; you can also visit the original URL.
The file type is application/pdf.
We distinguish three types: waveform coding, vocoding (or source coding) and hybrid coding. ...
In our simulations, background noise is simulated by adding samples of a rnemoryless Gaussian noise source with zero mean and variance a to the original speech samples. ...
This option can be useful for coding pur poses. ...
doi:10.1016/0165-1684(85)90136-7
fatcat:cdh3wpjqpncy5k7hbx77dflyuu
Data Analysis for Precision 21 cm Cosmology
[article]
2019
arXiv
pre-print
Preserved Fulltext
A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2020; you can also visit the original URL.
The file type is application/pdf.
Other Versions
on careful hardware design and telescope construction. ...
The redshifted 21 cm line is an emerging tool in cosmology, in principle permitting three-dimensional surveys of our Universe that reach unprecedentedly large volumes, previously inaccessible length scales ...
In deriving the normalizing scalar in Equation (143), we made use of the flat-sky approximation. ...
arXiv:1907.08211v2
fatcat:zjiv4owkqngr7az4cduoele4by
Properties of minimally doubled fermions
[article]
2017
arXiv
pre-print
Preserved Fulltext
A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2020; you can also visit the original URL.
The file type is application/pdf.
Hence, restoration of broken symmetries in the continuum limit requires three counterterms, one of which is power-law divergent. ...
Two independent non-perturbative procedures for removal of the power-law divergence are developed by means of a numerical study of hadronic observables for one variant of MDF in quenched approximation. ...
The simulation code is based on the Kpipi Code [73] [74] [75] , which was originally designed for simulations of overlap fermions in the quenched approximation. ...
arXiv:1706.07104v1
fatcat:w7zksxlk2jdq3hsrs2ulfo3xqe
An Introduction to Neural Networks
[chapter]
2015
The Mathematics of Banking and Finance
Preserved Fulltext
A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2017; you can also visit the original URL.
The file type is application/pdf.
Depending on the application, one can choose to perform the vector quantisation before learning the function approximation, or one can choose to learn the quantisation and the approximation layer simultaneously ...
These blocks can then be coded separately, stored or transmitted, where after a reconstruction of the whole image can bemade based on these coded 8 8 blocks. ...
doi:10.1002/9781119205920.ch31
fatcat:lzg3lko7c5ftbfkttkxaravxxi
An introduction to neural networks
1998
ChoiceReviews
Preserved Fulltext
A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2017; you can also visit the original URL.
The file type is application/pdf.
Depending on the application, one can choose to perform the vector quantisation before learning the function approximation, or one can choose to learn the quantisation and the approximation layer simultaneously ...
These blocks can then be coded separately, stored or transmitted, where after a reconstruction of the whole image can bemade based on these coded 8 8 blocks. ...
doi:10.5860/choice.35-5700
fatcat:ggsglohg4fa2zgh264ierzk6q4
« Previous
Showing results 1 — 15 out of 125 results