Segmental SNR Research Articles

Analysis of statistical estimators and neural network approaches for speech enhancement

Speech communicated is adversely affected by environmental noise. It is important to process the speech and reduce noise for better understanding. Speech enhancement or noise reduction is useful to provide comfort for human or machine listening. Traditional algorithms provide better noise reduction and better-quality speech. Due to the non-stationary nature of noise and the quasi-stationary nature of speech, the traditional methods are proven inadequate in achieving high-quality speech. Later statistical estimators based on Gaussian, and super-Gaussian Probability Density Function (PDF) assumption further improved the enhancement performance. But still, non-stationary noise nature introduces artifacts in processed signal and results in decreased performance. It is observed that neural network approaches and the factorization approach provide better performance even under non-stationary noises by proper training and large database. Different features result in variations in output performance under unseen noise and speaker conditions. It is important to understand the importance and advantages of traditional methods, statistical estimators, and neural network approaches performances. To select the suitable method for a required application, it is essential to consider the trade-off between quality and distortion. In this work, the importance of speech enhancement methods is discussed. Performance measures used for understanding the speech enhancement like Signal to Noise Ratio (SNR), Segmental SNR, Log-Likelihood Ratio (LLR), Weighted Spectral Slope (WSS), Perceptual Evaluation of Speech Quality (PESQ), Short-Time Objective Intelligibility (STOI), Signal to Distortion Ratio (SDR) and Mean Opinion Score (MOS), are given. Highlights of important results are discussed for analyzing better speech enhancement methods for the required application. In this work, performance is compared using objective and subjective performance measures. Simulation results show superior performance when neural network is employed in statistical estimators.

Security analysis of encrypted audio based on elliptic curve and hybrid chaotic maps within GFDM modulator in 5G networks

Wireless communications face significant security challenges, so there is an ongoing necessity to develop an appropriate security strategy to protect data from eavesdroppers using cryptography based on chaos theory. Generalized frequency division multiplexing (GFDM) is a modern multicarrier waveform adaptable to 5G requirements, but its security issues have not been considered. Therefore, this paper proposed an efficient security technique within the GFDM modulator to protect the audio transmission against eavesdropping in 5G networks. The proposed GFDM is achieved by separating the subsymbols for the subcarrier into real and imaginary parts and combining them using a mixture of elliptic curve-linear congruential generator sequence (EC-LCG) with Ikeda and Tent maps, respectively. After that, the subsymbols of each subcarrier are permuted independently using the Duffing map. The effectiveness of the proposed approach to resist attacks was tested, and findings that were achieved are histogram, signal to noise ratio (SNR=-27.8068), spectral segment SNR (SSSNR=-34.9912), peak SNR (PSNR=0.9142), frequency weighted log spectral distance (dFWLOG=37.498), cepstral distance (dCD=9.0176), mean square error (MSE=0.82097), keyspace, and speed. These results show that the proposed model provides a high-security level, high speed in the encryption/decryption, large keyspace, and high sensitivity to the initial conditions.

A robust speech enhancement method in noisy environments

Speech enhancement aims to eliminate or reduce undesirable noises and distortions, this processing should keep features of the speech to enhance the quality and intelligibility of degraded speech signals. In this study, we investigated a combined approach using single-frequency filtering (SFF) and a modified spectral subtraction method to enhance single-channel speech. The SFF method involves dividing the speech signal into uniform subband envelopes, and then performing spectral over-subtraction on each envelope. A smoothing parameter, determined by the a-posteriori signal-to-noise ratio (SNR), is used to estimate and update the noise without the need for explicitly detecting silence. To evaluate the performance of our algorithm, we employed objective measures such as segmental SNR (segSNR), extended short-term objective intelligibility (ESTOI), and perceptual evaluation of speech quality (PESQ). We tested our algorithm with various types of noise at different SNR levels and achieved results ranging from 4.24 to 15.41 for segSNR, 0.57 to 0.97 for ESTOI, and 2.18 to 4.45 for PESQ. Compared to other standard and existing speech enhancement methods, our algorithm produces better results and performs well in reducing undesirable noises.

Single-channel speech enhancement by PSO-GSA with harmonic regeneration noise reduction

Speech quality significantly affects the performance of speech dependent systems. Noise in the background lowers the clarity and intelligibility of speech. The augmentation of speech can increase its quality. We propose a single-channel speech improvement framework that combines particle swarm optimization (PSO), gravitational search algorithm (GSA), and harmonic regeneration noise reduction (HRNR) to minimize speech signal noise and increase speech intelligibility. The proposed hybrid algorithm optimizes the amount of overlap between the noisy speech frames. This helps in reducing the overlapped noise. Then HRNR algorithm is applied to retain the speech harmonics. The algorithm gives improvement in the speech intelligibility for babble, car and exhibition noise. The segmental signal to noise ratio (SNR) is also improved for these noise types. There is improvement in speech intelligibility with minimal speech distortion.

Curvelet based robust improved sine adaptive filter for feedback cancellation in hearing aids

Acoustic feedback is a common and persistent issue in hearing aids, which can lead to limitations in the achievable amplification and significant degradation in sound quality, including howling artefacts. Recently, the maximum versoria criteria (MVC) algorithm has been developed to reduce the acoustic feedback in hearing aids in the presence of impulsive noise. However, the robust technique MVC has not addressed the feedback signal’s sparseness characteristic while developing an adaptive feedback canceller. This paper proposes a robust and sparsity-aware technique called curvelet-improved sine-based adaptive filtering (CISAF) algorithm for adaptive feedback cancellation (AFC) in hearing aids using the prediction error method (PEM) in the case of non-gaussian interference. The step factor is enhanced based on the logarithmic cosh function to speed up the algorithm’s convergence while guaranteeing a low steady-state error. This variable step size (VSS) is included with CISAF as CISAF-VSS technique. Computer simulations show that the suggested method achieves a good convergence rate, maintains high speech quality, keeps a high segmental SNR (segSNR) and short-time objective intelligibility (STOI) at steady-state.

On DCT-based MMSE estimation of short time spectral amplitude for single-channel speech enhancement

This paper proposes Discrete Cosine Transform (DCT) based speech enhancement algorithms. These algorithms utilize minimum mean square error (MMSE) estimator of clean short-time spectral amplitude, which respectively uses Gaussian, Laplace and Gamma probability density functions (PDF) as speech priors. We consider the noise process is additive and Gaussian. The proposed estimators are closed-form solutions, whereas the conventional Discrete Fourier Transform (DFT) based estimators derived under super-Gaussian speech priors have no closed-form solutions. We also examine the estimators with the Speech Presence Uncertainty (SPU) that addresses the speech or silence problem with probability. Compared to the alternative approaches, such as the Ephraim and Malah or the Erkelens et.al MMSE-STSA estimators, the proposed methods demonstrate superior performance in terms of Segmental SNR (SegSNR), Perceptual Evaluation of Speech Quality (PESQ), short-time objective intelligibility measure (STOI), and mean subjective preference score, while exhibiting an equal or lower complexity.

Impact of University Classroom Size on the Relationship between Speech Quality and Intelligibility

In this paper, five objective measures of the quality of speech signals distorted by reverberation are compared with the Speech Transmission Index (STI). The main aim of the comparison is to further test and explain the reasons for the previously discovered phenomenon of an increase in the speech quality and intelligibility with increasing room size. The comparison is performed for three university classrooms of small, medium and large sizes. The correlation coefficients between the quality and intelligibility estimates of speech obtained for 5-6 points of each room are estimated. Speech signal quality is assessed using intrusive measures such as segmental signal-to-noise ratio (SSNR), log-spectral distortion (LSD), frequency-weighted segmental signal-to-noise ratio (FWSNR), bark spectral distortion (BSD), and perceptual evaluation of speech quality (PESQ). For BSD, high correlation coefficients (0.57-0.99) are determined for rooms of all sizes and an increase in the correlation coefficient with the room size increase is found, which can be explained by a decrease in the density of early sound reflections. For FWSNR, high correlation (0.65-0.98) is determined for medium and large rooms. For PESQ, high correlation (0.96-0.99) is obtained for large classroom. SSNR and LSD are found to be uncorrelated with STI for rooms of all sizes.

A speech separation system in video sequence using dilated inception network and U-Net

In this paper, an audio-visual model for separating a speech of the target speaker from a combination of other speakers’ speeches is proposed. It can be used in speech separation, automatic speech recognition systems (ASR) and also in creating single speaker speech databases. Speech separation is complicated problem using audio information only so visual and auditory signals are combined to complete the separation process. The proposed model consists of four modules, two for audio signal, one for visual feature and the last one used to concatenate the features resulted from the previous three modules to generate the separated signals. Our proposed model improved Short-time objective intelligibility (STOI) with 11%, Perceptual Evaluation of Speech Quality (PESQ) with 24%, and Frequency-weighted Segmental SNR (fwSNRseg) with 16% compared with previous works. It also improved Csig' which is the predicted rating of speech distortion with 13% and 'Covl' which is the predicted rating of overall quality with 18% compared with previous audio-visual models.

Noise Reduction of Lung Sounds based on Singular Spectrum Analysis combined with Discrete Cosine Transform

Lung sound signals are sensitive to environmental noise. Much research has been conducted on the enhancement of pulmonary sound. This study investigated the normal lung sounds as bronchovesicular (BV) and vesicular (V) signals and proposed a novel denoising method called SSA-DCT. Using Singular Spectrum Analysis (SSA), the noise-related components were separated from respiratory information components. An algorithm was also proposed to determine the information and noise time interval, and by applying Discrete Cosine Transform (DCT), the signal energy was attenuated in the noise intervals. Moreover, the concept of safety component, which is secure against noise, was introduced. Then, an algorithm for automatic identification of BV and V signals using the safety component was presented. The error of this algorithm at SNR of 10 dB was 5%. Lung sounds were recorded from 12 healthy subjects using four channels over the posterior chest wall. The signals were recorded in an acoustic laboratory and then contaminated with additive white Gaussian noise with different levels of SNR. The respiratory signals were also recorded in a relatively quiet environment with real ambient noise and denoised by the proposed method. The proposed method was compared with Coiflet wavelet decomposition with hard SureShrink thresholding. The denoising performance of both methods was evaluated using qualitative and quantitative approaches. The SSA-DCT method (e.g., at an SNR level of 10 dB) with average segmental SNR improvements of 2.52 and 3.44 dB for the BV and V signals, respectively, is significantly superior to the wavelet analysis with average segmental SNR improvements of 0.89 and 1.53 dB for the BV and V signals, respectively.

An improved two-stage dereverberation method based on bayesian estimation of a speech source

It is not easy to extract speech emanating from a specific direction in a large reverberant room with additional sound sources. Many of the existing methods work well only if there is a single sound source, and their performance degrades if additional sound sources are present. In this work, we have addressed this problem by maximizing the posterior probability of speech signal as computed by use of Bayes' Theorem. Our de-reverb method is two-staged. We have evaluated the efficacy of our method against several popular methods in terms of five objective measures; Signal-to-Interference Ratio, Log-Likelihood Ratio, Cepstrum Distance, Frequency-Weighted Segment SNR, and Speechto-Reverberation Modulation Energy Ratio. We show that on each of these parameters, the method proposed in this work performs better than the one we have benchmarked it against.

SINGLE CHANNEL SPEECH ENHANCEMENT USING EVOLUTIONARY ALGORITHM WITH LOG-MMSE

Additive noise degrades speech quality and intelligibility. Speech enhancement reduces this noise to make speech more pleasant and intelligible. It plays a significant role in speech recognition or speech-operated systems. In this paper, we propose a single-channel speech enhancement method in which the log-minimum mean square error method (log-MMSE) and modified accelerated particle swarm optimization algorithm are used to design a filter for improving the quality and intelligibility of noisy speech. Accelerated particle swarm optimization (APSO) algorithm is modified in which a single dimension of particle position is changed in a single iteration while obtaining the particle’s new position. Using this algorithm, a filter is designed with multiple passbands and notches for speech enhancement. The modified algorithm converges faster compared with standard particle swarm optimization algorithm (PSO) and APSO giving optimum filter coefficients. The designed filter is used to enhance the speech. The proposed speech enhancement method improves the perceptual estimation of speech quality (PESQ) by 17.05% for 5dB babble noise, 33.92 % for 5dB car noise, 14.96 % for 5dB airport noise, and 39.13 % for 5dB exhibition noise. The average output PESQ for these four types of noise is improved compared to conventional methods of speech enhancement. There is an average of 7.58 dB improvement in segmental SNR for these noise types. The proposed method improves speech intelligibility with minimum speech distortion.

Audio Signal Extraction and Enhancement Based on CNN From Laser Speckles

A micro vibration signal extraction method based on deep neural network is proposed. Rough surface of vibrating object modulates the illuminating laser wave front and generates speckle pattern, which is recorded by a linear array CMOS and preprocessed and input into a 16-layer convolution neural network (CNN) trained with specially prepared data. The optical experimental setup is analyzed to fulfil the temporal and spatial Shannon sampling theorem. The output audio signals are evaluated with standard algorithms and show enhanced segmental SNR and intelligibility. The effects of different input audio types and quality of raw audio signals are investigated, and the results show that the neural network is robust to the input. The CNN structure is optimized and the results show the performance decrease with the reduction of convolution layers. The performances of three popular deep neural networks are compared and the performance of CNN is better.

A New Speech Enhancement Technique Based on Stationary Bionic Wavelet Transform and MMSE Estimate of Spectral Amplitude

Speech enhancement has gained considerable attention in the employment of speech transmission via the communication channel, speaker identification, speech-based biometric systems, video conference, hearing aids, mobile phones, voice conversion, microphones, and so on. The background noise processing is needed for designing a successful speech enhancement system. In this work, a new speech enhancement technique based on Stationary Bionic Wavelet Transform (SBWT) and Minimum Mean Square Error (MMSE) Estimate of Spectral Amplitude is proposed. This technique consists at the first step in applying the SBWT to the noisy speech signal, in order to obtain eight noisy wavelet coefficients. The denoising of each of those coefficients is performed through the application of the denoising method based on MMSE Estimate of Spectral Amplitude. The SBWT inverse, S B W T − 1 , is applied to the obtained denoised stationary wavelet coefficients for finally obtaining the enhanced speech signal. The proposed technique’s performance is proved by the calculation of the Signal to Noise Ratio (SNR), the Segmental SNR (SSNR), and the Perceptual Evaluation of Speech Quality (PESQ).

Real-time high sensibility vibration detection based on phase correlation of line speckle patterns

A vibration extraction system based on laser speckle with phase correlation method is presented. The speckle pattern sequences are recorded by a linear array CMOS, which allows the system to detect high frequency signals more than 20 kHz. The line speckle patterns are fast Fourier transformed and correlated between the adjacent rows in the frequency domain. The phase angles of correlation calculations are directly used as the displacements between the respective two rows in the object space. The recovered audio signals are evaluated by segmental SNR (SegSNR), normalized sub-band envelope correlation (NSEC), and log-likelihood ratio (LLR). Compared with the centroid and quadratic fitting method, the phase correlation method obtains advantages of high stability, high sensibility, and real-time.

SUPPRESSION OF WHITE NOISE FROM THE MIXTURE OF SPEECH AND IMAGE FOR QUALITY ENHANCEMENT

This study proposed a correlation analysis of two recent approaches. The FAST ICA technique is used for the separation of the multimodal data (i.e, mixture of audio, noise and image signal) and the minimum mean-square error (MMSE) is used for the removal of white noise from the audio signal. Initially, multimodal data will be formed by combining all the three signals (i.e. a mixture of audio, noise and image signals). For creating an ideal situation and for SNR comparisons, separation of the signals will be performed using the Fast ICA technique. ICA, Independent element analysis is a recently developed technique in which the goal is to seek a linear interpretation of non-Gaussian knowledge for the elements to be as statistically free as possible. Such representations record the key structure of the data in several applications, including signal quality and signal separation. ICA learns a linear decay of the data. ICA can find the basic elements and sources included in the data found where traditional methods fail. After the separation of the mixed data, denoising will be performed using the MMSE technique. The main purpose of the MMSE technique is to remove White Noise from the unmixed audio signal which will be further used for overall and segmental SNR comparisons for quality enhancement. Based on the designed algorithms, both of these techniques are real-time data-driven programs. These techniques are explored with standard De-noising methods using several different estimation methods like signal-to-noise ratio (SNR). Experimental results prove that the proposed MMSE technique works well for both noise segmentation and overall consideration of noise distortion signals. These statistical techniques can be used in many applications, such as in different communication systems to eliminate background noise and in channels to reduce channel interference between different applications in speech communications

Low-complexity artificial noise suppression methods for deep learning-based speech enhancement algorithms

Deep learning-based speech enhancement algorithms have shown their powerful ability in removing both stationary and non-stationary noise components from noisy speech observations. But they often introduce artificial residual noise, especially when the training target does not contain the phase information, e.g., ideal ratio mask, or the clean speech magnitude and its variations. It is well-known that once the power of the residual noise components exceeds the noise masking threshold of the human auditory system, the perceptual speech quality may degrade. One intuitive way is to further suppress the residual noise components by a postprocessing scheme. However, the highly non-stationary nature of this kind of residual noise makes the noise power spectral density (PSD) estimation a challenging problem. To solve this problem, the paper proposes three strategies to estimate the noise PSD frame by frame, and then the residual noise can be removed effectively by applying a gain function based on the decision-directed approach. The objective measurement results show that the proposed postfiltering strategies outperform the conventional postfilter in terms of segmental signal-to-noise ratio (SNR) as well as speech quality improvement. Moreover, the AB subjective listening test shows that the preference percentages of the proposed strategies are over 60%.

Performance measurement of a hybrid speech enhancement technique

For enhancing the performance of the quality of speech, this paper presents a novel, hybrid speech enhancement technique based on the combination of two-step noise reduction (TSNR), harmonic regeneration noise reduction (HRNR), and a comb filter. This technique outperforms existing methods that are based on TSNR, HRNR, Wavelet, and hybrid TSNR-HRNR. The parameters such as spectrogram, average segmental SNR (ASSNR), mean square error (MSE), mean opinion score (MOS), perceptual evaluation of speech quality (PESQ), and diagnostic rhyme test (DRT) were obtained for the performance comparison. Results revealed that the proposed method outperforms in terms of ASSNR, MSE, speech quality, and intelligibility when compared to the TSNR, HRNR, wavelet, and TSNR-HRNR based speech enhancement methods.

Perceptually weighted β-order spectral amplitude Bayesian estimator for phase compensated speech enhancement

A speech enhancement (SE) technique improves the quality and intelligibility of the noise degraded speech. The conventional MMSE-based SE technique minimizes the Bayesian cost function by modifying the magnitude-spectra, whereas keeping the phase-spectra intact. In this current study, both magnitude and phase spectra are modified in order to enhance the noisy speech using a phase compensated perceptually weighted β-order Bayesian estimator. In the proposed SE technique, we first modified the phase of noisy speech spectra alone. Secondly, we used perceptually motivated β-order Bayesian estimator to modify the magnitude-spectra. The estimator holds the advantages of the perceptually-weighted and the β-order spectral amplitude estimator to get a better gain function. The compensated phase spectra and the estimated magnitude spectra are combined to reconstruct the noise attenuated enhanced speech signals. The proposed speech enhancement technique is evaluated for different noise levels (0 dB to +10 dB) in terms of the objective quality and intelligibility measures using the NOIZEUS and AURORA databases. It is observed that the proposed enhancement technique considerably improves the quality and maintains intelligibility in non-stationary and stationary noisy environments. The experimental results show that the average short-time objective intelligibility (STOI), normalized covariance metric (NCM), perceptual evaluation of speech quality (PESQ), segmental SNR (SegSNR), and Likelihood-Ratio (LLR) are improved by 12.2%, 9.25%, 19.69%, 31.25%, and 39.06% over the noisy unprocessed speech signals. The proposed technique also outscored the competing techniques.

Hybrid PSO-NLMS (HPSO-NLMS) algorithm for blind speech quality enhancement in time domain

Speech intelligibility enhancement by adaptive two-channel filtering algorithms in the time domain is a fundamental problem in the areas of signal processing application. In this paper, we address this problem and suggest a new adaptive hybrid algorithm combined with Forward blind source separation (FBSS) structure to retrieve a speech signal from noisy observations. Usually, the Normalized Least Mean Square algorithm (NLMS) is widely used in adaptive filtering application, however conventional NLMS presents difficulties to solve the trade-off between the speed of convergence and the low steady state error. Recently, numerous variable step-size NLMS (VNLMS) algorithms were proposed to solve this trade-off, our proposed algorithm falls within the same type of algorithm. The proposed Hybrid PSO-NLMS (HPSO-NLMS) algorithm is a combination of the NLMS and the Particle Swarm Optimization algorithms (PSO), where we use PSO to obtain the appropriate NMLS step size at each iteration. We integrate the new HPSO-NLMS in the FBSS structure, and compare their performances to conventional NLMS and PSO algorithms. We show that the proposed algorithm can produce better noise reduction performance, in terms of several objective metric such as the system misalignment (SM), the Segmental signal to noise Ratio (SegSNR) and the segmented mean square error (SegMSE).

Performance analysis of neural network, NMF and statistical approaches for speech enhancement

Bayesian Estimators are very useful in speech enhancement and noise reduction. But, it is noted that the traditional estimators process only amplitudes and the phase is left unprocessed. Among the Bayesian estimators, Super- Gaussian based estimators provide improved noise reduction. Super-Gaussian Bayesian estimators, which uses processed phase information for estimation of amplitudes provides further improved results. In this work, the Complex speech coefficients given Uncertain Phase (CUP) based Bayesian estimators like CUP-GG (CUP Estimator with speech spectral coefficients assumed as Gamma and noise spectral coefficients as Generalized Gamma), CUP-NG (Speech as Nakagami) are compared under white noise, pink noise, Babble noise and Non-Stationary factory noise conditions. The statistical estimators show less effective results under completely non-stationary assumptions like non-stationary factory noise, babble noise etc. Non-negative Matrix Factorization (NMF) based algorithms show better performance for non stationary noises. The drawback of NMF is, it requires apriori knowledge about speech. This drawback can be overcome by taking the advantages of both statistical approaches and NMF approaches. NR-NMF and WR-NMF speech enhancement methods are developed by providing posteriori regularization based on statistical assumption of speech and noise DFT coefficients distribution. Also a speech enhancement method which uses CUP-GG estimator and NMF with online noise bases update are considered for comparison. The progress in neural network based approaches for speech enhancement further shown that with large dataset and better training, the speech enhancement algorithms results in improved results. In this work, the neural network approach for speech enhancement is implemented and compared the method with traditional estimators and NMF approaches. For generalization of unseen noise types the proposed neural network approach uses dropout. Also for training the network, the features obtained from apriori SNR and aposteriori SNR is used in this method. The objective of this paper is to analyze the performance of speech enhancement methods based on Neural Network, NMF and statistical based. The objective performance measures Perceptual Evaluation of Speech Quality (PESQ), Short-Time Objective Intelligibility (STOI), Signal to Noise Ratio (SNR), Segmental SNR (Seg SNR) are considered for comparison.