Microphone array speech enhancement based on optimized IMCRA

2021 ◽  
Vol 69 (6) ◽  
pp. 468-476
Author(s):  
Qiuying Li ◽  
Tao Zhang ◽  
Yanzhang Geng ◽  
Zhen Gao
Keyword(s):  
Power Spectrum ◽  
Speech Enhancement ◽  
Signal To Noise Ratio ◽  
Microphone Array ◽  
Noise Power ◽  
Noise Power Spectrum ◽  
Signal To Noise ◽  
Noise Ratio ◽  
Speech Segments ◽  
Enhancement Algorithm

Microphone array speech enhancement algorithm uses temporal and spatial informa- tion to improve the performance of speech noise reduction significantly. By combining noise estimation algorithm with microphone array speech enhancement, the accuracy of noise estimation is improved, and the computation is reduced. In traditional noise es- timation algorithms, the noise power spectrum is not updated in the presence of speech, which leads to the delay and deviation of noise spectrum estimation. An optimized im- proved minimum controlled recursion average speech enhancement algorithm, based on a microphone matrix is proposed in this paper. It consists of three parts. The first part is the preprocessing, divided into two branches: the upper branch enhances the speech signal, and the lower branch gets the noise. The second part is the optimized improved minimum controlled recursive averaging. The noise power spectrum is updated not only in the non-speech segments but also in the speech segments. Fi- nally, according to the estimated noise power spectrum, the minimum mean-square error log-spectral amplitude algorithm is used to enhance speech. Testing data are from TIMIT and Noisex-92 databases. Short-time objective intelligibility and seg- mental signal-to-noise ratio are chosen as evaluation metrics. Experimental results show that the proposed speech enhancement algorithm can improve the segmental signal-to-noise ratio and short-time objective intelligibility for various noise types at different signal-to-noise ratio levels.

Power spectral analysis of the diaphragm electromyogram

1983 ◽  
Vol 54 (6) ◽  
pp. 1579-1584 ◽  
Author(s):  
T. K. Aldrich ◽  
J. M. Adams ◽  
N. S. Arora ◽  
D. F. Rochester
Keyword(s):  
Power Spectrum ◽  
Signal To Noise Ratio ◽  
Normal Subjects ◽  
Chronic Obstructive ◽  
Signal To Noise ◽  
Low Frequencies ◽  
Mean Values ◽  
The Mean ◽  
Noise Ratio ◽  
Diaphragm Fatigue

We studied the power spectrum of the diaphragm electromyogram (EMG) at frequencies between 31 and 246 Hz in four young normal subjects and five patients with chronic obstructive lung disease (COPD). Diaphragm EMGs were analyzed during spontaneous breathing and maximum inspiratory efforts to determine the effect of signal-to-noise ratio on the power spectrum and if treadmill exercise to dyspnea was associated with diaphragm fatigue. We found that the centroid frequencies of the power spectra (fc) were strongly correlated (r = 0.93) with ratios of power at high frequencies to power at low frequencies (H/L) for all subjects. Of the two indices, H/L had the largest standard deviation expressed as a percentage of the mean. The mean values of both of these decreased significantly after exercise, fc from 100.2 to 97.3 and H/L from 1.07 to 0.97. Signal-to-noise ratios were higher in maximal inspiratory efforts and after exercise in normal subjects and higher in COPD patients. The signal-to-noise ratio was correlated negatively with fc and H/L, indicating that these indices of the shape of the power spectrum are influenced by signal strength and noise levels as well as muscle function. We conclude that the fc and H/L index similar qualities of the power spectrum, that they are partially determined by the signal-to-noise ratio, and that, in some cases, exercise to dyspnea is associated with apparently mild diaphragm fatigue.

The Noise Analysis of Fluid Systemsin Solid-State Nanopore Sensors

2013 ◽  
Vol 419 ◽  
pp. 517-520 ◽  
Author(s):  
Song Ying ◽  
Lei Wang ◽  
Wen Yuan Zhao
Keyword(s):  
Solid State ◽  
Signal To Noise Ratio ◽  
Noise Analysis ◽  
Power Spectra ◽  
Noise Power ◽  
Label Free ◽  
Signal To Noise ◽  
Fluid Systems ◽  
Noise Ratio ◽  
Nanopore Sensors

The solid-state nanopore sensor offers a versatile platform for the rapid, label-free electrical detection and analysis of single molecules, especially on DNA sequencing. However, the overall signal-to-noise ratio (SNA) is a major challenge in sequencing applications. In our work, two different fluid systems made by metal and plexiglass have been designed to improve the signal to noise ratio of the solid-state nanopore sensor. From the measurements on the noise power spectra with a variety of conditions, it is found that plexiglass fluid system coupled with shielding box produces a good quality of electric signals on nanopore sensors.

A NEW TECHNIQUE FOR THE STUDY OF SCATTER PROPAGATION IN THE TROPOSPHERE

1957 ◽  
Vol 35 (8) ◽  
pp. 823-830 ◽  
Author(s):  
J. H. Chapman ◽  
W. J. Heikkila ◽  
J. E. Hogarth
Keyword(s):  
Power Spectrum ◽  
Communication System ◽  
Carrier Frequency ◽  
Signal To Noise Ratio ◽  
Signal Strength ◽  
Noise Power ◽  
Signal To Noise ◽  
A New Technique ◽  
Conventional Signal ◽  
Noise Power Density

The power spectrum of the fluctuations in received signal strength on a near-optical U.H.F. circuit has been measured. The sidebands associated with these fluctuations can overlap the information-carrying sidebands of a communication system. When this happens, these sidebands must be taken into account in determining the signal-to-noise ratio of the system. In other words, the fluctuations then have the characteristics of noise, and therefore they are called propagation noise in the present paper. Experiments at a carrier frequency of 500 Mc. have shown that the propagation noise power density usually varies with sideband frequency ƒ (measured from the carrier) as 1/ƒ2, for f in the range 0.1 to 10 c.p.s. Departures from this law have been observed in the regions near 0.1 c.p.s. and 10 c.p.s. The measurement of the power spectrum directly offers several advantages over the conventional signal strength recording method, and these are discussed herein.

scholarly journals Design of an Indigenised Electronic Lock-In Amplifier

2000 ◽  
Vol 23 (1) ◽  
pp. 53-60
Author(s):  
Umesh Kumar
Keyword(s):  
Narrow Band ◽  
Signal To Noise Ratio ◽  
Broad Band ◽  
Power Line ◽  
Frequency Drift ◽  
Noise Power ◽  
Signal To Noise ◽  
Band Noise ◽  
Noise Ratio ◽  
Lock In

An indigenised lock-in amplifier is designed that enables the accurate measurement of signals contaminated by broad-band noise, power-line pick-up, frequency drift, or other sources of interference. It does this by means of an extremely narrow band detector which has the centre of its passband locked to the frequency of the signal to be measured. Large improvements in signal-to-noise ratio are achieved.

scholarly journals The Signal to Noise Ratio in Speckle Interferometry

1979 ◽  
Vol 50 ◽  
pp. 23-1-23-18 ◽  
Author(s):  
J.C. Dainty ◽  
A.H. Greenaway
Keyword(s):  
Power Spectrum ◽  
Signal To Noise Ratio ◽  
Binary Star ◽  
Image Plane ◽  
Light Level ◽  
Speckle Interferometry ◽  
Signal To Noise ◽  
Speckle Image ◽  
Auto Correlation Function ◽  
Noise Ratio

AbstractRecent theoretical studies of the signal to noise ratio (SNR) of photon limited speckle (image plane) interferometry are reviewed. The SNR of an estimate of the object power spectrum is evaluated for both the single and double aperture cases, for arbitrary light levels. The SNR for the auto-correlation function method of analysis is also given for the low light level case and applied to the special case of binary star observations. The SNRs for the power spectrum and autocorrelation function analyses are compared and a comparison is also made between speckle (image plane) and amplitude (pupil or aperture plane) interferometry. Limiting observable magnitudes are estimated for some relevant cases.

Sign in / Sign up

Export Citation Format

Share Document