Suppression Of Additive Noise Research Articles

Microphone Array Beamforming With High Flexible Interference Attenuation and Noise Reduction

This paper studies the problem of microphone array beamforming to enhance a speech signal of interest in adverse acoustic environments, where interference and additive background noise coexist. The problem is formulated as one of convex optimization whose solution under a specified level of interference attenuation leads to an interference controlled maximum noise reduction (ICMR) beamformer, which can be expressed as a linear combination of two MVDR beamformers: one attempts to extract the desired source signal while the other attempts to extract the interference. The combination coefficients are functions of the array manifold vectors, noise coherence matrix, and the specified interference attenuation factor. By tuning the interference attenuation factor, the ICMR beamformer can be implemented to achieve aggressive interference attenuation or even eliminate interference completely; but this may lead to less additive noise suppression or even noise amplification. To control the maximum sacrifice in gain (SG) of the signal-to-noise ratio (SNR) that is acceptable for additive reduction, a variant of ICMR is derived, which is named as the ICMR-SG beamformer. Simulations are performed and the results show that the ICMR beamformer is able to control the amount of interference attenuation. In comparison, ICMR-SG controls the maximum SG of SNR while achieving the optimal possible level of interference attenuation.

A Hybrid Deep-Learning Approach for Single Channel HF-SSB Speech Enhancement

The high-frequency single side-band (HF-SSB) speech radio is an essential technique for long-distance speech transmission. However, the HF-SSB received speech is corrupted by both high-power noise and severe channel fading, and the typical speech enhancement methods only focus on the suppression of additive noise. In this letter, a two-stage hybrid approach is proposed to enhance the speech quality of the HF-SSB radio. In the anti-fading stage, we adopt the anti-fading convolution neural network (AF-CNN) to eliminate the effects of channel fading. In the noise suppression stage, noise suppression CNN (NS-CNN) subnet and unsupervised denoising block are used parallelly to further improve the performance and the generalization ability. In addition, we present the optimal topological relations of noise suppression and anti-fading modules by comparison and analysis. Experimental results show that applying the AF-CNN subnet before noise suppression can effectively help recover the weak speech components. Moreover, in terms of objective intelligibility and quality scores, the overall performance of the proposed method outperforms the typical methods that only consider noise suppression.

Sparse Approach in Filtering of Color Images Corrupted by Mixture Noises

A novel filtering approach is presented in denoising in the color images contaminated by mixture of additive-impulsive noises. Novel framework consists of three principal stages: impulsive noise suppression that is performed detecting pixels corrupted by impulsive noise and then, filtering found spikes by a variant of median filter; during second stage, original additive noise suppression filter is employed in Wavelet transform domain via a sparse representation and 3D-filtering; finally, nondesirable effects obtained in an image during previous stages are processed to correct fine details. In case of multiplicative noise suppression, the designed denoising scheme uses 3D homomorphic sparse processing stage and post-filtering procedure. Evaluation of novel approach in denoising complex distortions has been performed using objective criteria (PSNR and SSIM measures) and subjective perception via human visual system confirming their better performance in comparison with state-of-theart techniques.

External serrodyne modulation for the suppression of low-frequency noise in quadrature interferometry.

In interferometry, reaching a high signal-to-noise ratio at low frequencies can be challenging when the additive noise is nonstationary. Although this problem is typically solved by inserting a frequency shifter into one of the arms, in some cases, the interferometer cannot or should not be modified in this way. This Letter presents an alternative solution, based on external serrodyne frequency modulation, which is comparable to the typical approach in terms of complexity and performance yet does not require the modification of a passive interferometer. We demonstrate a prototype that achieves frequency shifting at 500 kHz with 89% power efficiency, leading to the wideband suppression of low-frequency additive noise by more than 19 dB. This enables a fully passive measurement of the thermoconductive noise of a 100 m single-mode fiber.

The Speech Enhancement via Attention Masking Network (SEAMNET): An End-to-end System for Joint Suppression of Noise and Reverberation

This paper proposes the Speech Enhancement via Attention Masking Network (SEAMNET), a neural network-based end-to-end single-channel speech enhancement system designed for joint suppression of noise and reverberation. It formalizes an end-to-end network architecture, referred to as b-Net , which accomplishes noise suppression through attention masking in a learned embedding space. A key contribution of SEAMNET is that the b-Net architecture contains both an enhancement and an autoencoder path. This paper proposes a novel loss function which simultaneously trains both the enhancement and the autoencoder paths, so that disabling the masking mechanism during inference causes SEAMNET to reconstruct the input speech signal. This allows dynamic control of the level of suppression applied by SEAMNET via a minimum gain level, which is not possible in other state-of-the-art approaches to end-to-end speech enhancement. This paper also proposes a perceptually-motivated waveform distance measure. In addition to the b-Net architecture, this paper proposes a novel method for designing target waveforms for network training, so that joint suppression of additive noise and reverberation can be performed by an end-to-end enhancement system, which has not been previously possible. Experimental results show the SEAMNET system to outperform a variety of state-of-the-art baselines systems, both in terms of objective speech quality measures and subjective listening tests. Finally, this paper draws parallels between SEAMNET and conventional statistical model-based enhancement approaches, offering interpretability of many network components.

Novel Proposal for Prediction of CO2 Course and Occupancy Recognition in Intelligent Buildings within IoT

Many direct and indirect methods, processes, and sensors available on the market today are used to monitor the occupancy of selected Intelligent Building (IB) premises and the living activities of IB residents. By recognizing the occupancy of individual spaces in IB, IB can be optimally automated in conjunction with energy savings. This article proposes a novel method of indirect occupancy monitoring using CO2, temperature, and relative humidity measured by means of standard operating measurements using the KNX (Konnex (standard EN 50090, ISO/IEC 14543)) technology to monitor laboratory room occupancy in an intelligent building within the Internet of Things (IoT). The article further describes the design and creation of a Software (SW) tool for ensuring connectivity of the KNX technology and the IoT IBM Watson platform in real-time for storing and visualization of the values measured using a Message Queuing Telemetry Transport (MQTT) protocol and data storage into a CouchDB type database. As part of the proposed occupancy determination method, the prediction of the course of CO2 concentration from the measured temperature and relative humidity values were performed using mathematical methods of Linear Regression, Neural Networks, and Random Tree (using IBM SPSS Modeler) with an accuracy higher than 90%. To increase the accuracy of the prediction, the application of suppression of additive noise from the CO2 signal predicted by CO2 using the Least mean squares (LMS) algorithm in adaptive filtering (AF) method was used within the newly designed method. In selected experiments, the prediction accuracy with LMS adaptive filtration was better than 95%.

Method of Image Denoising in Generalized Phase Space with Improved Indicator of Spatial Resolution

The paper proposes a nonlocal method of additive noise suppression in digital image based on presenting the image in matrix phase space and using nonconventional methods of multivariate statistical analysis, namely, the surrogate data technology that makes it possible to generate a pseudo-ensemble of surrogate images with their subsequent averaging using a single snapshot. This approach is based on properties of the coherent accumulation of signal component of the observation and noncoherent accumulation of its noise component as the size of observation ensemble increases that allows us to partially solve the contradiction between the denoising level and the distortion or loss of small-sized details of image, i.e. reduction of spatial resolution. The simulation modeling of the proposed method of generalized SDT-filtering of noise was conducted using the application software package of MathCad and Matlab. A comparative analysis of the spatial resolution of the proposed and several known methods of denoising has been carried out using the resolution–measurement criterion and the modified Rayleigh criterion. As is shown, the proposed method demonstrates a better spatial resolution as compared to the most common methods of denoising that is confirmed by the results of simulation modeling.

Free Space Optic Receiver with Strongly Overlapped Photodetectors’ Field of View

In this study, we designed a mobile free space optic receiver that uses several photodetectors to provide omnidirectional receiving capability. Assuming only one transmitter, it is a receiver which builds a single input multiple output optical channel. The photodetectors are fixed to truncated pyramid walls. Electrical signals from the photodetectors are processed using an equal gain combining technique. This architecture allows simple circuits and enables additive noise suppression. The minimum angle between the pyramid base and the direction of falling rays was calculated to determine the threshold for additive noise suppression. Two areas of interest presented themselves: the processing of very weak electrical signals often drowned in noise, and optimization of the number of photodetectors whose fields of view overlapped strongly. We outline the design of the optical receiver circuitry and provide some practical hints concerning its assembly. The receiver was evaluated using bit error rate measurements and comparing signal-to-noise ratio parameters for various photodetector numbers. The measured data confirm the theoretical assumptions.

Sparse Technique for Images Corrupted by Mixed Gaussian-Impulsive Noise

In this paper, a novel framework is presented for denoising images that have been corrupted by a mixture of additive and impulsive noise. The proposed method consists of three main stages: impulsive noise suppression, additive noise suppression and post-processing. In the first stage, a pixel that has been contaminated by impulsive noise is detected and filtered. In the next stage, filtering is based on sparse representation and 3D-processing using discrete cosine transform. Finally, the post-processing stage increases the filtering quality by using a bilateral filter and an edge restoration technique. Evaluation is performed using objective criteria (PSNR and SSIM) and subjective human visual perception to confirm the methods performance compared with state-of-the-art techniques.

Algorithms for Estimating Signal Information Parameters under the Action of Broadband Non-Gaussian Noise

This paper describes the synthesis of algorithms for estimating signal information parameters with noninertial nonlinear transformation of the input mixture of the signal and broadband non-Gaussian noise. There is also the analysis of algorithms optimal in the domain of small mismatch and quasioptimal algorithms, which are capable of operating in arbitrary mismatches between measured parameters and their estimates in the case of random signal/noise ratio at the gauge input. This work also presents the obtained basic estimated relations that help one determine the form of optimal amplitude characteristics of the nonlinear transformation unit, the quality of suppression of additive noise in a given unit with optimal and arbitrary amplitude characteristic, the steepness of the discrimination characteristic, and the value of the phase characteristic in the domain of small mismatches. It is shown that, by specifying the type of action and the transfer function of a smoothing chain, it is possible to determine the steady-state dynamic and fluctuation errors in a closed servo-system.

Note: A new method for directly reducing the sampling jitter noise of the digital phasemeter.

The sampling jitter noise is one non-negligible noise source of the digital phasemeter used for space gravitational wave detection missions. This note provides a new method for directly reducing the sampling jitter noise of the digital phasemeter, by adding a dedicated signal of which the frequency, amplitude, and initial phase should be pre-set. In contrast to the phase correction using the pilot-tone in the work of Burnett, Gerberding et al., Liang et al., Ales et al., Gerberding et al., and Ware et al. [M.Sc. thesis, Luleå University of Technology, 2010; Classical Quantum Gravity 30, 235029 (2013); Rev. Sci. Instrum. 86, 016106 (2015); Rev. Sci. Instrum. 86, 084502 (2015); Rev. Sci. Instrum. 86, 074501 (2015); and Proceedings of the Earth Science Technology Conference (NASA, USA, 2006)], the new method is intrinsically additive noise suppression. The experiment results validate that the new method directly reduces the sampling jitter noise without data post-processing and provides the same phase measurement noise level (10-6 rad/Hz1/2 at 0.1 Hz) as the pilot-tone correction.

MuLoG, or How to Apply Gaussian Denoisers to Multi-Channel SAR Speckle Reduction?

Speckle reduction is a longstanding topic in synthetic aperture radar (SAR) imaging. Since most current and planned SAR imaging satellites operate in polarimetric, interferometric, or tomographic modes, SAR images are multi-channel and speckle reduction techniques must jointly process all channels to recover polarimetric and interferometric information. The distinctive nature of SAR signal (complex-valued, corrupted by multiplicative fluctuations) calls for the development of specialized methods for speckle reduction. Image denoising is a very active topic in image processing with a wide variety of approaches and many denoising algorithms available, almost always designed for additive Gaussian noise suppression. This paper proposes a general scheme, called MuLoG (MUlti-channel LOgarithm with Gaussian denoising), to include such Gaussian denoisers within a multi-channel SAR speckle reduction technique. A new family of speckle reduction algorithms can thus be obtained, benefiting from the ongoing progress in Gaussian denoising, and offering several speckle reduction results often displaying method-specific artifacts that can be dismissed by comparison between results.

Additive and exclusive noise suppression by iterative trimmed and truncated mean algorithm

An iterative trimmed and truncated arithmetic mean (ITTM) algorithm is proposed, and the ITTM filters are developed. Here, trimming a sample means removing it and truncating a sample is to replace its value by a threshold. Simultaneously trimming and truncating enable the proposed filters to attenuate the mixed additive and exclusive noise in an effective way. The proposed trimming and truncating rules ensure that the output of the ITTM filter converges to the median. It offers an efficient method to estimate the median without time-consuming data sorting. Theoretical analysis shows that the ITTM filter of size n has a linear computational complexity O(n). Compared to the median filter and the iterative truncated arithmetic mean (ITM) filter, the proposed ITTM filter suppresses noise more effectively in some cases and has lower computational complexity. Experiments on synthetic data and real images verify the filter's properties.

Correlation-Based and Model-Based Blind Single-Channel Late-Reverberation Suppression in Noisy Time-Varying Acoustical Environments

This paper considers suppression of late reverberation and additive noise in single-channel speech recordings. The reverberation introduces long-term correlation in the observed signal. In the first part of this work, we show how this correlation can be used to estimate the late reverberant spectral variance (LRSV) without having to assume a specific model for the room impulse responses (RIRs) while no explicit estimates of RIR model parameters are needed. That makes this <i xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">correlation-based</i> approach more robust against RIR modeling errors. However, the correlation-based method can follow only slow time variations in the RIRs. Existing <i xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">model-based</i> methods use statistical models for the RIRs, that depend on one or more parameters that have to be estimated blindly. The common statistical models lead to simple expressions for the LRSV that depend on past values of the spectral variance of the reverberant, noise-free, signal. All existing model-based LRSV estimators in the literature are derived assuming the RIRs to be <i xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">time-invariant realizations</i> of a stochastic process. In the second part of this paper, we go one step further and analyze <i xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">time-varying</i> RIRs. We show that in this case the reverberance tends to become decorrelated. We discuss the relations between different RIR models and their corresponding LRSV estimators. We show theoretically that similar simple estimators exist as in the time-invariant case, provided that the reverberation time <i xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">T</i> <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">60</sub> and direct-to-reverberation ratio (DRR) of the RIRs remain nearly constant during an interval of the order of a few frames. We show that the reverberation time can be taken frequency-bin independent in DFT-based enhancement algorithms. Experiments with time-varying RIRs validate the analysis. Experiments with additive nonstationary noise and time-invariant RIRs show the influence of blind estimation of the reverberation time and the DRR.

The concept of signal-to-noise ratio in the modulation domain and speech intelligibility

A new concept is proposed that relates to intelligibility of speech in noise. The concept combines traditional estimations of signal-to-noise ratios (S/N) with elements from the modulation transfer function model, which results in the definition of the signal-to-noise ratio in the modulation domain: the (SN)(mod). It is argued that this (SN)(mod), quantifying the strength of speech modulations relative to a floor of spurious modulations arising from the speech-noise interaction, is the key factor in relation to speech intelligibility. It is shown that, by using a specific test signal, the strength of these spurious modulations can be measured, allowing an estimation of the (SN)(mod) for various conditions of additive noise, noise suppression, and amplitude compression. By relating these results to intelligibility data for these same conditions, the relevance of the (SN)(mod) as the key factor underlying speech intelligibility is clearly illustrated. For instance, it is shown that the commonly observed limited effect of noise suppression on speech intelligibility is correctly "predicted" by the (SN)(mod), whereas traditional measures such as the speech transmission index, considering only the changes in the speech modulations, fall short in this respect. It is argued that (SN)(mod) may provide a relevant tool in the design of successful noise-suppression systems.

Digital, phase-sensitive detection for in situ diode-laser spectroscopy under rapidly changing transmission conditions

A new harmonic detection scheme for fully digital, fast-scanning wavelength-modulation spectroscopy (DFS-WMS) is presented. DFS-WMS is specially suited for in situ absorption measurements in combustion environments under fast fluctuating transmission conditions and is demonstrated for the first time by open-path monitoring of ambient oxygen using a distributed-feedback diode laser, which is doubly modulated with a fast linear 1 kHz-scan and a sinusoidal 300 kHz-modulation. After an analog high-pass filter, the detector signal is digitized with a 5 megasample/s 12-bit AD-converter card plugged into a PC and subsequently – unlike standard lock-ins – filtered further by co-adding 100 scans, to generate a narrowband comb filter. All further filtering and the demodulation are performed completely digitally on a PC with the help of discrete Fourier transforms (DFT). Both 1f- and 2f-signals, are simultaneously extracted from the detector signal using one ADC input channel. For the 2f-signal, a linearity of 2% and a minimum detectable absorption of 10-4 could be verified experimentally, with the sensitivity to date being limited only by insufficient gain on the 2f-frequency channel. Using the offset in the 1f signal as a transmission ‘probe’, we could show that the 2f-signal can be transmission-corrected by a simple division by the 1f-background, proving that DFS-WMS provides the possibility of compensating for transmission fluctuations. With the inherent suppression of additive noise, DFS-WMS seems well suited for quantitative in situ absorption spectroscopy in large combustion systems. This assumption is supported by the first measurements of oxygen in a high-pressure combustor at 12 bar.

Mechanisms of perceptual learning

Systematic measurements of perceptual learning were performed in the presence of external or stimulus noise. In the new external noise method (Dosher, B, & Lu, Z.-L. (1997). Investigative Ophthalmology and Visual Science, 38, S687; Lu, Z.-L., & Dosher, B. (1998). Vision Research, 38, 1183–1198), increasing amounts of external noise (white Gaussian random noise) is added to the visual stimulus in order to identify mechanisms of perceptual learning. Performance improved (threshold contrast was reduced) over days of practice on a peripheral orientation discrimination task—labelling Gabor patches as tilted slightly to the right or left. Practice improvements were largely specific to the trained quadrant of the display. Performance improved at all levels of external noise. The external noise method and perceptual template model (PTM) of the observer identifies the mechanism(s) of performance improvements as due to stimulus enhancement, external noise exclusion, or internal noise suppression. The external noise method was further extended by measuring thresholds at two threshold performance levels, allowing identification of mixtures in the PTM model. Perceptual learning over 8–10 days improved the filtering or exclusion of external noise by a factor of two or more, and improved suppression of additive internal noise—equivalent to stimulus enhancement—by 50% or more. Coupled improvements in external noise exclusion and stimulus enhancement in the PTM model may reflect channel weighting. Perceptual learning may not reflect neural plasticity at the level of basic visual channels, nor cognitive adjustments of strategy, but rather plasticity at an intermediate level of weighting inputs to decision.

Polyspectral analysis of mixed processes and coupled harmonics

Polyspectral analysis of processes with mixed spectra is considered, and scaled polyperiodograms are introduced to clarify issues related to stationarity, ergodicity, and suppression of additive stationary noise in harmonic retrieval problems. Spectral and polyspectral approaches are capable of retrieving (un)coupled harmonics, not only when the harmonics have constant amplitudes, but also when they are observed in nonzero mean multiplicative noise. Fourier series polyspectra and asymptotic properties of scaled polyperiodograms provide general tools for higher order analysis of time series with mixed spectra. A single record phase coupling detector is derived to obviate the assumption of independent multiple records required by existing methods. The novelties are illustrated by simulation examples.

Optical DPSK/Heterodyne‐. Detection scheme using two polarization planes

AbstractThis paper proposes the dual polarization DPSK (DPDPSK) as an extension of the polarization modulation to the narrowband system. the basic performances are examined. First, it is shown that the proposed scheme occupies half of the signal bandwidth of the polarization modulation scheme, the receiver configuration is the same as that of the polarization modulation, and the phase noise insensitivity, which is the feature of the polarization modulation, is utilized effectively in the proposed system.Assuming a receiver with the double‐filtering configuration, a theoretical analysis for the receiver error rate is presented, considering the envelope fluctuation due to the phase noise at the bandpass filter (BPF) output, as well as the additive noise suppression by the lowpass filter (LPF) in the demodulator. It is shown that the optimal BPF bandwidth exists that maximizes the receiver sensitivity.