PCA-based Compression Research Articles

A Synchrophasor Data Compression Technique With Iteration-Enhanced Phasor Principal Component Analysis

The phasor data were compressed as separated amplitudes and phases in previous synchrophasor data compression techniques. To utilize the spatial correlation and temporal continuity of synchrophasors for data compression, a phasor principal component analysis (PPCA) in the field of complex numbers is proposed to compress synchrophasors as a whole in this article. Then, an iterative phasor principal components selection method is proposed to achieve PPCA and ensure the accuracy of reconstructed data since the existing eigenvalue-based criteria are not suitable for data compressions. Moreover, the proposed PPCA is enhanced by an iteration-based process to reduce the computation of PPCA. Actual PMU data measured under both a low-frequency oscillation incident and a two-phase short circuit incident conditions are used to verify the performance of PPCA compared with a recent PCA-based compression method. The results demonstrate that PPCA achieves higher compression ratios with better accuracy of reconstructed data, significantly reduced computation, and better real-time performance under both conditions.

Spatio-temporal Segmentation Based Adaptive Compression of Dynamic Mesh Sequences

With the recent advances in data acquisition techniques, the compression of various dynamic mesh sequence data has become an important topic in the computer graphics community. In this article, we present a new spatio-temporal segmentation-based approach for the adaptive compression of the dynamic mesh sequences. Given an input dynamic mesh sequence, we first compute an initial temporal cut to obtain a small subsequence by detecting the temporal boundary of dynamic behavior. Then, we apply a two-stage vertex clustering on the resulting subsequence to classify the vertices into groups with optimal intra-affinities. After that, we design a temporal segmentation step based on the variations of the principal components within each vertex group prior to performing a PCA-based compression. Furthermore, we apply an extra step on the lossless compression of the PCA bases and coefficients to gain more storage saving. Our approach can adaptively determine the temporal and spatial segmentation boundaries to exploit both temporal and spatial redundancies. We have conducted extensive experiments on different types of 3D mesh animations with various segmentation configurations. Our comparative studies show the advantages of our approach for the compression of 3D mesh animations.

CSI Partitioning Method with PCA-Based Compression for Low-Complexity Feedback of Large-Dimensional Channels

This letter considers a codebook-based channel state information (CSI) feedback technique for large-dimensional channels, e.g., massive multiple-input multiple-output (MIMO) channels. We propose a CSI partitioning method, focus on the search complexity required in order to obtain the best beamforming vector, and investigate the tradeoff between the complexity and the performance. Then, we present an analytical framework that provides design guidelines for balancing the complexity , performance , and feedback overhead of practical large-scale MIMO systems.

A Novel Noise Removal Method Using Neural Networks

In this paper is presented a new technique consisting in applying some pre-whitening and shrinkage methods followed by a neural network-based supervised approach for correlated noise removal purposed. In our work the type of noise and the covariance matrix of noise are known or can be estimated using the “white wall” method. Due to data dimensionality, a PCA-based compression technique is used to obtain a tractable solution. The local memories of the neurons are determined using a supervised learning process based on the compressed pre-processed inputs and the compressed version of the original images. The proposed method is evaluated using some of the most commonly used indicators and the results are reported in the third section of the paper. The conclusive remarks together with suggestions for further work are supplied in the final part of the paper.

Channel Correlation Modeling and its Application to Massive MIMO Channel Feedback Reduction

In this paper, we propose a feedback information reduction technique for massive multiple-input multiple-output (MIMO) systems. To this end, we analytically derive a covariance matrix of spatially correlated Rayleigh fading channels in closed form. The covariance matrix is expressed based on its statistics, including transmit and receive antennas' correlation factors, channel variance, and channel delay profile. The closed-form expression enables a principal component analysis (PCA)-based compression of channel state information (CSI), which allows the feedback overhead to be efficiently reduced. We also analyze the compression feedback error, bit-error-rate (BER) performance, and the spectral efficiency (SE) of the system using the PCA-based compression. Under our proposed model, numerical results verify that the PCA-based compression method significantly reduces the feedback overhead of the massive MIMO systems with marginal performance degradation from full-CSI feedback. Furthermore, we propose a new design framework by numerically showing that there exists the optimal number of transmit antennas in terms of SE for a given limited feedback amount.

Compressive mining: fast and optimal data mining in the compressed domain

Real-world data typically contain repeated and periodic patterns. This suggests that they can be effectively represented and compressed using only a few coefficients of an appropriate basis (e.g., Fourier and wavelets). However, distance estimation when the data are represented using different sets of coefficients is still a largely unexplored area. This work studies the optimization problems related to obtaining the tightest lower/upper bound on Euclidean distances when each data object is potentially compressed using a different set of orthonormal coefficients. Our technique leads to tighter distance estimates, which translates into more accurate search, learning and mining operations directly in the compressed domain. We formulate the problem of estimating lower/upper distance bounds as an optimization problem. We establish the properties of optimal solutions and leverage the theoretical analysis to develop a fast algorithm to obtain an exact solution to the problem. The suggested solution provides the tightest estimation of the $$L_2$$ L 2 -norm or the correlation. We show that typical data analysis operations, such as $$k$$ k -nearest-neighbor search or k-Means clustering, can operate more accurately using the proposed compression and distance reconstruction technique. We compare it with many other prevalent compression and reconstruction techniques, including random projections and PCA-based techniques. We highlight a surprising result, namely that when the data are highly sparse in some basis, our technique may even outperform PCA-based compression. The contributions of this work are generic as our methodology is applicable to any sequential or high-dimensional data as well as to any orthogonal data transformation used for the underlying data compression scheme.

Reconstruction of total radiance spectra of fluorescent samples by means of nonlinear principal component analysis

Nonlinear principal component analysis (NLPCA) was used for compression and reconstruction of the total radiance factors (TRFs) of fluorescent samples. The spectral dataset included a total of 358 fluorescent reflectance spectra in the visible range of the spectrum. Spectral data compression was followed by extracting the parameterized nonlinear manifolds using the NLPCA technique. To compare the performance of NLPCA-based compression with the linear method, the orthonormal feature vectors of the dataset were also extracted by the linear PCA. The spectral performance of NLPCA and PCA-based compression approaches was assessed by the root mean square error and the goodness-fitting coefficient between the real and the reconstructed spectra. The percentages of feasible spectra by each method, i.e., those with nonnegative TRFs, were also reported as other criteria for the evaluation of methods. Furthermore, the colorimetric performance of methods were appraised by the measuring the CIELAB 1976 color difference values between the actual and reconstructed spectra under illuminants D65 and A and the 1964 standard observer. The NLPCA-based compression method performed significantly better than the standard PCA-based technique particularly in the lower dimensional spaces of the spectral radiance factors of fluorescent colors.

A psychophysically validated metric for bidirectional texture data reduction

Bidirectional Texture Functions (BTF) are commonly thought to provide the most realistic perceptual experience of materials from rendered images. The key to providing efficient compression of BTFs is the decision as to how much of the data should be preserved. We use psychophysical experiments to show that this decision depends critically upon the material concerned. Furthermore, we develop a BTF derived metric that enables us to automatically set a material's compression parameters in such a way as to provide users with a predefined perceptual quality. We investigate the correlation of three different BTF metrics with psychophysically derived data. Eight materials were presented to eleven naive observers who were asked to judge the perceived quality of BTF renderings as the amount of preserved data was varied. The metric showing the highest correlation with the thresholds set by the observers was the mean variance of individual BTF images. This metric was then used to automatically determine the material-specific compression parameters used in a vector quantisation scheme. The results were successfully validated in an experiment with six additional materials and eighteen observers. We show that using the psychophysically reduced BTF data significantly improves performance of a PCA-based compression method. On average, we were able to increase the compression ratios, and decrease processing times, by a factor of four without any differences being perceived.

Weighted compression of spectral color information

Spectral color information is used nowadays in many different applications. Accurate spectral images are usually very large files, but a proper compression method can reduce needed storage space remarkably with a minimum loss of information. In this paper we introduce a principal component analysis (PCA) -based compression method of spectral color information. In this approach spectral data is weighted with a proper weight function before forming the correlation matrix and calculating the eigenvector basis. First we give a general framework for how to use weight functions in compression of relevant color information. Then we compare the weighted compression method with the traditional PCA compression method by compressing and reconstructing the Munsell data set consisting of 1,269 reflectance spectra and the Pantone data set consisting of 922 reflectance spectra. Two different weight functions are proposed and tested. We show that weighting clearly improves retention of color information in the PCA-based compression process.

New hybrid opto-electronic method for fast and unsupervised object detection

We present an unsupervised method for the fast detection of noisy objects of varying sizes and contrasts, which is implementable on an opto-electronic device. This method consists of first achieving multiresolution representation of the noisy data using successive Gaussian filterings. Then, the resulting set of Gaussian smoothings is compressed using principal component analysis (PCA). This compression is applied within regions of interest (ROI) that are previously detected using a fast technique adapted to the features of analyzed data. The different objects of interest are finally segmented using a standard valley thresholding technique, which is locally applied within each ROI. An experimental evaluation using synthetic images underlines the robustness of this method and its ability to achieve unsupervised detection of strongly noisy objects. Theoritical and experimental estimations of the computing power of a high-speed optical correlator and of a specialized digital processor have shown that it is faster to compute: optically global Gaussian filterings, the PCA-based compression being digitally performed. Experiments have also shown the potential of the proposed method for the fast detection of liver tumors from computer tomography (CT)-scan images. The feasibility of its hybrid opto-electronic implementation is demonstrated using an experimental optical correlator.