Signal Compression Research Articles

The CogniAid trial. The impact of two hearing aid signal processing strategies on cognition

BackgroundUntreated hearing loss is a risk factor for age-related cognitive decline and hearing aids have been shown to slow cognitive decline in a population at risk for dementia. This double-blind multiple site randomized trial tested the hypothesis that for older adults with below-average cognition, a “Simple” hearing aid fitting strategy (based on linear amplification with output limiting compression signal processing) would improve hearing and cognition more than a “Standard” approach (adaptive compression-based processing).MethodsTwo hundred and fifty-six adults aged over 65 were screened for cognitive function using the NIH toolbox cognitive battery. Participants with below median age-adjusted fluid composite cognitive scores (<100) were eligible to participate (n = 104). Sixty-seven eligible participants proceeded to trial and were randomized 1:1 to a simple or standard hearing aid fitting. Participants in the Standard group were fitted with hearing aids matched to non-linear real-ear prescription targets (either NAL-NL1 or NL2), while participants in the Simple group were fitted with hearing aids matched to linear prescription targets (NAL-R). Participants and researchers not fitting the hearing aids were blinded to allocation.ResultsForty-eight participants completed assessments in 12 months. The Standard hearing aid group improved on measures of fluid cognition and hearing. There was a statistically significant difference in fluid cognition scores between groups. The fluid cognition composite score for participants receiving the Simple fitting changed by 3.5 points. Those with the Standard fitting improved by 10.3 points. Hearing outcomes for each group were improved by the same amount.ConclusionThis is the first study to show that hearing aid fitting strategies using markedly different signal processing result in significantly different cognitive outcomes after 12 months of use. The Standard fitting resulted in greater improvement in cognition than the Simpler fitting which was the opposite result to what had been hypothesized. The results reinforce findings indicating hearing aid benefits for the elderly and that they improve cognition.

Novel sensor developments for photon science at the MPG semiconductor laboratory

The world of photon science experiences significant advancements since the advent of synchrotron light sources with unprecedented brilliance, intensity and pulse repetition rates, with large implications on the detectors used for instrumentation. Here, an overview about the work on this field carried out at the semiconductor laboratory of the Max-Planck-Society (MPG HLL) is given. Main challenges are high dynamic range to resolve faint features at the fringes of scatter images as well as structures in bright peaks, and high bandwidth to fully exploit the fast timing capability of the source. A newly developed device to improve the signal-to-noise-ratio (SNR) at high bandwidths is the so-called MARTHA (Monolithic Array of Reach-Through Avalanche Photodiodes) structure, which integrates an array of APDs on a monolithic substrate. The reach-through architecture assures near 100% fill factor and allows implementing a thin entrance window with optimized quantum efficiency for low energy X-rays. The structures operate in proportional mode with adjustable gain, and can serve as a drop-in replacement for PAD detectors in hybrid pixel systems. A more sophisticated solution for low to medium frame rate applications with high contrast requirement are pnCCDs with high dynamic range in the pixel area featuring DEPFET based readout nodes with non-linear amplification (NLA). The high dynamic range mode has been demonstrated for pnCCD devices with a pixel size down to 75 μm2. Framerates of up to 1 kHz are possible for a 1 Megapixel detector. Small size prototypes of these structures have recently been manufactured. Modified DEPFET structures with build-in non-linear amplification are also used to implement active pixel detectors optimized for high dynamic range. Successfully prototyped for the DSSC sensors (DEPFET Sensor with Signal Compression) at the XFEL, these structures are increasingly being used in applications requiring high contrast and intensity, e.g., TEM imaging. Charge handling capability and output characteristics can be tailored to the requirements, as well as pixel geometry and size. The large intrinsic gain of the DEPFET provides excellent SNR even at fast timing. Pixels can be read with a speed of 100 ns, the resulting frame rate depends on the degree of readout parallelization.

Gated recurrent unit predictor model-based adaptive differential pulse code modulation speech decoder

Speech coding is a method to reduce the amount of data needs to represent speech signals by exploiting the statistical properties of the speech signal. Recently, in the speech coding process, a neural network prediction model has gained attention as the reconstruction process of a nonlinear and nonstationary speech signal. This study proposes a novel approach to improve speech coding performance by using a gated recurrent unit (GRU)-based adaptive differential pulse code modulation (ADPCM) system. This GRU predictor model is trained using a data set of speech samples from the DARPA TIMIT Acoustic-Phonetic Continuous Speech Corpus actual sample and the ADPCM fixed-predictor output speech sample. Our contribution lies in the development of an algorithm for training the GRU predictive model that can improve its performance in speech coding prediction and a new offline trained predictive model for speech decoder. The results indicate that the proposed system significantly improves the accuracy of speech prediction, demonstrating its potential for speech prediction applications. Overall, this work presents a unique application of the GRU predictive model with ADPCM decoding in speech signal compression, providing a promising approach for future research in this field.

Underwater optical signal detection system using diffuser-based lensless imaging.

We propose a diffuser-based lensless underwater optical signal detection system. The system consists of a lensless one-dimensional (1D) camera array equipped with random phase modulators for signal acquisition and one-dimensional integral imaging convolutional neural network (1DInImCNN) for signal classification. During the acquisition process, the encoded signal transmitted by a light-emitting diode passes through a turbid medium as well as partial occlusion. The 1D diffuser-based lensless camera array is used to capture the transmitted information. The captured pseudorandom patterns are then classified through the 1DInImCNN to output the desired signal. We compared our proposed underwater lensless optical signal detection system with an equivalent lens-based underwater optical signal detection system in terms of detection performance and computational cost. The results show that the former outperforms the latter. Moreover, we use dimensionality reduction on the lensless pattern and study their theoretical computational costs and detection performance. The results show that the detection performance of lensless systems does not suffer appreciably. This makes lensless systems a great candidate for low-cost compressive underwater optical imaging and signal detection.

Kompresija JPEG i BPG bez gubitka vizuelnih informacija na primeru baze KonJND-1k

Introduction/purpose: This paper presents the results of the research on visually lossless image compression which is of particular interest because it achieves a high degree of compression, while the visual quality of the image is not impaired, i.e., end users are very satisfied with the image quality. The analysis was carried out using the publicly available large-scale picture-wise KonJND-1k database which contains the results of subjective tests on JPEG and BPG compressed images. Methods: Thanks to the availability of images from the KonJND-1k database, the dependence of objective assessments of image quality on parameters that control the degree of compression of source signals (quality factor for JPEG and quantization parameter for BPG) is analyzed. The results of the visually lossless subjective tests are used for a deep analysis of the boundary and typical values of the parameters that control these two types of compression, as well as for the analysis of the corresponding values of the objective quality scores. Furthermore, reliable features for predicting the boundary between visually lossless and visually lossy compression have been identified. For that purpose, the degree of agreement between the predictions and the ground truth values of the peak signal-to-noise ratio (PSNR) and image representation in bits per pixel (bpp) is used. The visually lossless compression ratio is used to compare JPEG and BPG techniques. Results: It is shown that the boundary between visually lossless and visually lossy image compression is found in a wide range of PSNR values (about 20 dB for JPEG and 15 dB for BPG). The corresponding JPEG image compression quality factor values at this threshold also range widely from 31 to 79, with concentration between 40 and 45. For the BPG encoder, the values of the quantization parameter are grouped around 30, and the boundary values are 25 and 34. Furthermore, it is shown that this boundary can be reliably determined based on simple features derived from the original uncompressed image. Gradient-based features known as spatial frequency and spatial information proved to be the best predictors. The degree of agreement between the predictions obtained from these features with the ground truth values of PSNR and bpp in both types of compression is greater than 85%. A comparative analysis has showed that, using BPG compression, it is possible, on the average, to achieve a twice larger compression ratio of visually lossless compression than for JPEG (80 versus 40). Conclusion: Although a high degree of agreement is achieved between the predictions and the ground truth values of PSNR and bpp of the boundary between visually lossless and visually lossy compression, there is a need for the development of new prediction approaches, especially with the BPG technique, which through the compression ratio proved to be superior to the JPEG technique. The existing databases used for the analysis of visually lossless compression contain color images from the visible part of the electromagnetic spectrum. Considering the increasing use of images from the infrared part of the spectrum, there is a need to conduct similar tests in this spectral range.

Compression of Higher-Order Ambisonic Signals Using Directional Audio Coding

Compression of Higher-Order Ambisonic Signals Using Directional Audio Coding

Tensor decomposition-based compression and noise reduction of multichannel ECG signals

Tensor decomposition-based compression and noise reduction of multichannel ECG signals

The high/low frequency balance drives tactile perception of noisy vibrations.

Noisy vibrotactile signals transmitted during tactile explorations of an object provide precious information on the nature of its surface. Understanding the link between signal properties and how they are interpreted by the tactile sensory system remains challenging. In this paper, we investigated human perception of broadband, stationary vibrations recorded during exploration of textures and reproduced using a vibrotactile actuator. Since intensity is a well-established perceptual attribute, we here focused on the relevance of the spectral content. The stimuli were first equalized in perceived intensity and subsequently used to identify the most salient spectral features using dissimilarity estimations between pairs of successive vibration. Based on dimensionally reduced spectral representations, models of dissimilarity ratings showed that the balance between low and high frequencies was the most important cue. Formal validation of this result was achieved through a Mushra experiment, in which participants assessed the fidelity of resynthesized vibrations with various distorted frequency balances. These findings offer valuable insights into human vibrotactile perception and establish a computational framework for analyzing vibrations as humans do. Moreover, they pave the way for signal synthesis and compression based on sparse representations, holding significance for applications involving complex vibratory feedback.

Noisy One-Bit Compressed Sensing With Side-Information

We consider the problem of sparse signal reconstruction from noisy one-bit compressed measurements when the receiver has access to side-information (SI). We assume that compressed measurements are corrupted by additive white Gaussian noise before quantization and sign-flip error after quantization. A generalized approximate message passing-based method for signal reconstruction from noisy one-bit compressed measurements is proposed which is then extended for the case where the receiver has access to a signal that aids signal reconstruction, i.e., side-information. Two different scenarios of SI are considered-a) SI consisting of support information only, and b) SI consisting of support and amplitude information. SI is either a noisy version of the signal or a noisy estimate of the support of the signal. We develop reconstruction algorithms from one-bit measurements using noisy SI available at the receiver. Laplacian distribution and Bernoulli distribution are used to model the two types of noises which when applied to the signal and the support, yield the SI for the above two cases, respectively. The Expectation-Maximization algorithm is used to estimate the noise parameter using noisy one-bit compressed measurements and the SI. We show that one-bit compressed measurement-based signal reconstruction is quite sensitive to noise, and the reconstruction performance can be significantly improved by exploiting available side-information at the receiver.

Compressed Sensing Vs. Auto-Encoder: On the Perspective of Signal Compression and Restoration

Compressed Sensing Vs. Auto-Encoder: On the Perspective of Signal Compression and Restoration

Compressed Sensing Based on an Improved K-SVD for Vibration Signal Compression Reconstruction in Wireless Sensor Networks

Compressed Sensing Based on an Improved K-SVD for Vibration Signal Compression Reconstruction in Wireless Sensor Networks

Biomedical signal compression using deep learning based multi-task compressed sensing

<span>Real-time transmission of biomedical signals is immensely challenging and requires cloud and internet of things (IoT) infrastructure. Security is also an important factor; however, to accomplish this, a reconstruction method is developed in which the entire signal is supplied as an input, the primary portion is considered here, and the signal is further encoded and transmitted to the destination. Electrocardiogram (ECG) compression for the lightweight wireless network is quite challenging for long-term healthcare monitoring. Compressed sensing (CS) involves efficient encoding mechanisms for error rate estimation for reconstruction and energy consumption for wireless transmission of data. We propose a multi-task compressed sensing (MT-CS) reconstruction mechanism in this study for ECG compression of data is most chosen for a wireless network system that has various sensors embedded in it. This model further extracts the essential adaptive features for correlation existing in the ECG signals. The performance of the proposed MT-CS reconstruction mechanism is evaluated on the multiparameter intelligent monitoring in intensive care (MIMIC-II) dataset, which ensures its robustness and generalization. The results obtained upon simulation ensure that the proposed MT-CS based reconstruction approach ensures excellent reconstruction signal with fewer measurements in comparison with the existing state-of-art CS techniques.</span>

Graph-Based EEG Signal Compression for Human–Machine Interaction

Graph-Based EEG Signal Compression for Human–Machine Interaction

An energy efficient reconfigurable architecture for multi-lead ECG signal compression

An energy efficient reconfigurable architecture for multi-lead ECG signal compression

Adaptive bias control of optical IQ modulator with low LFM dither and strong fluctuation resistance.

An innovative approach has been proposed for adaptive bias control in optical IQ modulators. In contrast to traditional approaches that utilize sine dither, this method employs a linear frequency modulated (LFM) signal as the dither, associated with the fractional Fourier Transform (FrFT) to extract the bias point drift. The LFM signal, after undergoing FrFT, transforms into a compressed signal (CS) with energy concentration in the fractional domain. Utilizing this signal for bias point monitoring, the proposed method demonstrates robust bias control even in the presence of substantial interferences, as substantiated by comprehensive simulations and experimental investigations. Remarkably, in a 20-Gbaud 16QAM signal transmission, the proposed approach achieves stable control of the bias point for over 4 hours, even in the presence of voltage fluctuations, while effectively reducing the dither amplitude by half. Furthermore, it maintains a low bit error rate (BER) below 10-5 even under intentional external interference.

Interference fading suppression for distributed acoustic sensor using frequency-shifted delay loop

In order to suppress the interference fading of phase-sensitive optical time domain reflectometer (Φ-OTDR), a multi-frequency detection method based on frequency-shifted delay loop (FSDL) is proposed. A probe pulse train with up to five consistent frequency components are generated by periodic frequency-shifted, amplification, delay and filtering processing in FSDL. The piecewise aggregate approximation (PAA) is introduced to compress the amount of operational data. Wavelet energy spectrum analysis and support vector machine (SVM) are used to extract features and realize the fading classification. With the help of fading label making and restoration, the multi-frequency signals are intelligently aggregated using rotated-vector-sum (RVS). Experimental results show that PAA has applicability in data compression of beat signals. After 5-layer wavelet energy spectrum analysis, the fading binary classification accuracy can reach 96.77 %, and the fading signals identified after segmentation can be restored to the original data. The fading probability based on SVM output labels can be reduced to 1.89 %. The SVM-based classification output labels aggregation shows that the vibration positioning signal-to-noise ratio (SNR) can be increased to 13.52 dB, the phase demodulation curve is smoother, and the demodulation SNR can be up to 17.63 dB. Finally, the R-Square of 0.997 for the amplitude of demodulated vibration signal in repeatability experiment verifies the validity of this effective scheme for fading recognition and suppression.

Radio Receiver with Internal Compression of Input Signals Using a Dispersive Delay Line with Bandpass Filters

Radio Receiver with Internal Compression of Input Signals Using a Dispersive Delay Line with Bandpass Filters

A likelihood based method for compressive signal recovery under Gaussian and saturation noise

Practical sensing systems have a limited dynamic range, which gives rise to saturation noise in signal acquisition if the signal attains values outside this range at certain locations. Despite rapid advances in compressed sensing theory and systems over the years, the effect of saturation on compressed sensing measurements, especially in the presence of additive Gaussian noise, has often been overlooked in many existing approaches. Some methods for compressive recovery under saturation noise simply discard saturated measurements, while others impose some ad hoc hard constraints to ensure consistency with the known saturation threshold. In this paper, we propose a novel maximum likelihood based estimator for compressive reconstruction in the presence of both Gaussian and saturation noise. Our proposed method ensures probabilistic consistency between the estimated signal and the saturated measurements and helps account for both the saturation effect and the potentially unbounded Gaussian noise in the measurements. We also derive upper bounds on the reconstruction error for our estimator, and argue that they follow intuitive trends. We analyze an important curvature term in these bounds, and show that it is superior to methods such as saturation rejection, indicating better stability and robustness. Furthermore, we present extensive simulation results to demonstrate the effectiveness of our method over several other existing methods in reconstructing synthetic signals and images from compressive measurements with saturation and Gaussian noise. We also show simulation results in audio signal de-clipping.

Latent ODEs-based data compression method for power system

In order to effectively compress the power system data and reduce the space occupied by data storage and transmission, this paper proposes a data compression method based on Latent Segmented ODE (LatSegODE) for the large-scale steady-state data generated during the working process of the power system. By training the ODE model using a large amount of steady-state data, a latent ODE model is generated to characterize the steady-state data of the power system, thereby significantly reducing the storage space occupied by the data. During the experimental stage, compression testing is carried out on the static data of the power system of a power grid company. According to the quantitative evaluation results of the two indices of normalized mean square error and data compression ratio, the proposed method has obvious compression advantages, and the compressed signals have a high degree of fitting to the actual sampled signal waveforms.

A REAL STRUCTURE-PRESERVING ALGORITHM FOR THE LOW- RANK DECOMPOSITION OF PURE IMAGINARY QUATERNION MATRICES AND ITS APPLICATIONS IN SIGNAL PROCESSING

With the increasing use of quaternions in fields such as quantum mechanics, rigid body rotation, signal and color image processing, and aerospace engineering, three- dimensional signal models represented by pure imaginary quaternion matrices have been cre- ated. This model treats the 3D information as a whole, and the processing preserves the intrinsic connections between the different channels. However, the computational process of quaternion algebra often generates real parts, which are inevitable. How to ensure the pure imaginary properties of 3D signal models is also a research priority. In this paper, we inves- tigate the low-rank decomposition of pure imaginary quaternion matrices using least squares iteration and give a real structure-preserving algorithm for the low-rank decomposition based on the isomorphic real representation. In addition, this matrix decomposition algorithm is applied to color image compression and 3D wave signal denoising problems. Numerical ex- periments show the effectiveness of the algorithm in this paper.