Spectral Subtraction Based Weighted Function for Pitch Extraction in Noisy Speech

Introduction: In order to assess how well conventional and hybrid pitch detection techniques perform in speech processing applications, a comparative analysis of the two types of methods is conducted. Method: A proposed hybrid approach, Proposed PEF+CEP, is examined alongside five traditional algorithms, namely Normalized Correlation Function (NCF), Pitch Estimation Filter (PEF), Log- Harmonic Summation (LHS), Summation of Residual Harmonics (SRH) and Cepstrum Pitch Determination (CEP). The effectiveness is evaluated using performance metrics like accuracy, specificity, sensitivity, and Gross Pitch Error (GPE). Results and Discussion: Our findings show that the accuracy and specificity of the traditional methods are impressive; the accuracy and sensitivity of the suggested hybrid method surpass their performance, with an astounding 98.8% accuracy and 99.2% sensitivity. Conclusion: Furthermore, the Proposed PEF+CEP method is a promising solution for accurate and dependable pitch detection in speech processing applications because it strikes a strong balance between computational efficiency and robustness. These results open up new avenues for research in the field of speech processing and demonstrate the potential of hybrid approaches.

Pitch is one of the most essential features in human speech analysis. Although numerous pitch detection methods have been developed, it is still a challenge to provide a high pitch detection performance in noisy environments. In this paper, we propose an anti-noise pitch detection method that combines a speech enhancement algorithm with a spectral flattening algorithm. In the experiments, we compare the proposed method with several widely used or state-of-the-art pitch detection methods. The results show that the proposed method has the lowest gross pitch error (GPE) rate among all the methods when dealing with white-noise added male speeches. Moreover, comparing the pitches estimated by the proposed method to those estimated by the conventional lag window method, we can see that the speech enhancement algorithm helps diminish pitch errors.

The complexity of the speech enhancement problem has motivated many different solutions. However, most techniques address situations in which the target speech is fully intelligible and the background noise energy is low in comparison with that of the speech. Thus while current enhancement algorithms can improve the perceived quality, the intelligibility of the speech is not increased significantly and may even be reduced. Recent research shows that intelligibility of very noisy speech can be improved by the use of a binary mask, in which a binary weight is applied to each time-frequency bin of the input spectrogram. There are several alternative goals for the binary mask estimator, based either on the Signalto-Noise Ratio (SNR) of each time-frequency bin or on the speech signal characteristics alone. Our approach to the binary mask estimation problem aims to preserve the important speech cues independently of the noise present by identifying time-frequency regions that contain significant speech energy. The speech power spectrum varies greatly for different types of speech sound. The energy of voiced speech sounds is concentrated in the harmonics of the fundamental frequency while that of unvoiced sounds is, in contrast, distributed across a broad range of frequencies. To identify the presence of speech energy in a noisy speech signal we have therefore developed two detection algorithms. The first is a robust algorithm that identifies voiced speech segments and estimates their fundamental frequency. The second detects the presence of sibilants and estimates their energy distribution. In addition, we have developed a robust algorithm to estimate the active level of the speech. The outputs of these algorithms are combined with other features estimated from the noisy speech to form the input to a classifier which estimates a mask that accurately reflects the time-frequency distribution of speech energy even at low SNR levels. We evaluate a mask-based speech enhancer on a range of speech and noise signals and demonstrate a consistent increase in an objective intelligibility measure with respect to noisy speech.

The fundamental frequency (F 0) represents pitch in speech that determines prosodic characteristics of speech and is needed in various tasks for speech analysis and synthesis.Despite decades of research on this topic, F 0 estimation at low signal-to-noise ratios (SNRs) in unexpected noise conditions remains difficult.This work proposes a new approach to noise robust F 0 estimation using a recurrent neural network (RNN) trained in a supervised manner.Recent studies employ deep neural networks (DNNs) for F 0 tracking as a frame-by-frame classification task into quantised frequency states but we propose waveform-tosinusoid regression instead to achieve both noise robustness and accurate estimation with increased frequency resolution.Experimental results with PTDB-TUG corpus contaminated by additive noise (NOISEX-92) demonstrate that the proposed method improves gross pitch error (GPE) rate and fine pitch error (FPE) by more than 35 % at SNRs between -10 dB and +10 dB compared with well-known noise robust F 0 tracker, PEFAC.Furthermore, the proposed method also outperforms state-of-the-art DNN-based approaches by more than 15 % in terms of both FPE and GPE rate over the preceding SNR range.

An Iterative Graph Spectral Subtraction Method for Speech Enhancement

The purpose of this paper is to compare the performance of two common filters operating on noisy speech recorded in automobiles travelling at various speeds. The filters are based on Spectral Subtraction (SS) and Kalman Filtering (KF). The literature contains studies based on simulated data whereas this paper uses real time data collected in car's in search of an optimal solution. The comparisons were based on real recorded samples containing noisy speech signals with durations of approximately 2 minutes each. Different cases of noise levels which represent the most common situations experienced by drivers were created. The different settings used include varying car speeds (e.g., 40 mph, 70 mph), varying fan power, and window positions settings. The study was carried out using three different car models. The measured noisy voice signals were filtered using the different filtering techniques and the resulting filtered signals were compared in the time domain and the frequency domain, both quantitatively and psychometrically. Furthermore, the quantitative analysis approach was applied to the results for more accurate interpretation. Results show that SS outperforms KF in noise reduction, and with much less speech distortion at the different Signal to Noise Ratios (SNRs) tested. The audio test results subjected to human listening are comparable with the simulation results. Overall, SS showed superior performance over KF in vehicular hands-free speech applications.

The paper proposes a pitch detection algorithm based on the short-time average magnitude difference function (AMDF) and the short-term autocorrelation function (ACF). At first, AMDF values are computed by AMDF algorithm for a frame of speech signal. And then ACF values are computed by ACF algorithm for the AMDF values. In order to decreases computational expense and complexity, the AMDF values of the frame of speech signal are then transformed into one bit signals. The method can also decrease the effects of amplitude and formants the speech signal for pitch detection. The pitch period is calculated by ACF algorithm for the one bit signals. The multiplication operation for short-time autocorrelation function of the one bit signals is replaced by simple addition operation. The algorithm can largely reduce its computational cost compared with calculating ACF directly. The experiments show that the algorithm can decreases computational expense and complexity compared with ACF algorithm. The algorithm can decrease the error rate of pitch detection compared with AMDF algorithm. The proposed algorithm is easy to implement in real-time signal processing

Environmental noise, known as an additive noise, not only corrupts the spectra of a speech signal but also blurs the shape of its energy contour. The corruption of the energy contour can distort the energy derived feature and degrade the pattern classification performance of noisy speech. To reduce the distortion of the energy feature, the energy bias in the energy contour has to be removed before the feature extraction. For this purpose, we propose two methods to estimate the noise energy; one is obtained from the speech inactive period, and one is from the noisy speech itself. The methods are evaluated by the connected digit recognition of TIDigits, in which the test speech is corrupted with white noise, babble, factory noise, and in-car noises. As shown in the experiments, the energy enhancement can provide an additional improvement when it is jointly applied with a spectral subtraction.

This paper describes a method for noise-proof detection of the fundamental frequency of the voice in a noisy environment. Noise reduction techniques have been required in the development of a hearing aid, because noise makes intelligibility of hearing awfully inferior. In various methods of noise reduction, the fundamental frequency is often a significant parameter, but it is difficult to extract the frequency from the noisy voice. In order to utilize a comb filter method for noise reduction, a new method of detecting the fundamental frequency is developed by using the property of continuity in the fundamental frequency and the power spectrum envelope (PSE) of the human voice. The continuity of the PSE is utilized for determining the most reliable frequency. The gross pitch error (GPE) is reduced by the determination. Besides the frequency used for the comb filter is obtained from a linear predicting frequency and the latest fundamental frequency from the noisy voice, so as to suppress fluctuation of the frequency that degrades filtered voice. The procedure improves a fine pitch error (FPE) within 5%. The results of the evaluation showed that the present method proved to be superior to the traditional cepstrum method in the GPE and the FPE. We conclude that the proposed frequency detection method is available for noise reduction in the comb filter method.

Speech is the audible acoustic signal generated by the articulatory system (lungs, vocal folds, vocal tract, tongue, lips) to communicate language. The fundamental frequency (F_0) is the lowest frequency component of a speech waveform and corresponds to the vibration rate of the vocal folds during voiced speech. It also determines the pitch of the speaker’s voice. In speech signal processing, this acoustic waveform is captured and analyzed to extract information about what is being said, how it is being said and who is saying it. In various speech processing applications such as voice synthesis, speaker recognition and emotion analysis accurate extraction of the fundamental frequency (F_0) is a vital task. However, the vocal tract’s formants can sometimes significantly alter the glottal waveform’s shape, making it difficult to identify the true pitch. Additionally, in the presence of background noise, traditional pitch detection techniques often experience a considerable decline in performance. This work proposes a robust method for extracting the fundamental frequency by utilizing the complementary advantages of the power spectrum and logarithmic spectrum in noisy speech environments. The power spectrum mitigates noise effects, while the logarithmic operation effectively separates vocal tract characteristics from the source excitation. The proposed approach integrates autocorrelation-based power spectral analysis with cepstral techniques derived from the log spectrum to improve pitch estimation accuracy under adverse conditions. Experimental results on noisy speech datasets show that the proposed hybrid method achieves lower gross pitch error and greater robustness than traditional methods such as BaNa, autocorrelation and cepstral techniques.

As with many approaches to noise robust automatic speech recognition (ASR) the benefits of spectral subtraction tend to diminish as noise levels in the order of 0 dB are approached. Whilst the majority of related work focuses on reducing magnitude errors a number of new approaches addressing the often overlooked, additional sources of error have appeared in the literature in recent years. Relatively lacking in the literature, however, is an empirical assessment which compares the effects of each error when noisy speech is processed by spectral subtraction. Such studies are vital in order to appreciate the potential penalty in performance when sources of error are overlooked. The objective in this paper is to assess, through ASR, the performance penalty associated with each source of error when noisy speech is treated with spectral subtraction. Experimental evidence based on two standard European databases and ASR protocols illustrates that, perhaps contrary to popular belief, for noise levels in the order of 0 dB and below, these often overlooked sources of error can lead to non-negligible degradations in performance. Whilst not a new idea, here the original emphasis is a thorough assessment that empirically highlights both the fundamental limitations and potential benefit of including the full complement of errors in the spectral subtraction model.

Speech enhancement is of paramount importance in improving the intelligibility and audibility of speech signals in an uncontrolled noisy environment over wireless communication systems. However, the existing techniques used to achieve this such as spectral subtraction are characterized by remnant noise or musical note, while Kalman filter which can remove the remnant noise is associated with instability and non-linear characteristics. Therefore, this paper proposes speech enhancement model in noisy environment using hybrid spectral-Kalman filter over wireless communication system. The proposed model was carried out using acquired noisy speech signals by the Samsung Android mobile phone. The phone call was made in an uncontrolled noisy environment. The signal received was processed by the hybrid spectral-Kalman filter which was achieved by using the output speech signal from the spectral subtraction in time domain as the input of the Kalman filter. The proposed hybrid spectral-Kalman filtering technique was evaluated using Signal-to-Noise Ratio (SNR), Mean Square Error (MSE), Perceptual Evaluation of Speech Quality (PESQ) and Short-Time Objective Intelligibility (STOI). Comparison was then performed with the existing techniques and validation was performed using corrupted speech signals obtained from the NOIZEUS corpus data set. The result revealed that the proposed technique performed better than the existing techniques in a noisy environment.

In this paper, we propose a confidence measure for the pitch estimation method presented in a parallel paper [1]. The confidence measure is derived based on a sparse representation of the speech harmonic structure. The measurement quantitatively reflects the harmonicity associated with an estimated pitch. It is used to indicate the reliability of the results. Histogram is employed to illustrate the distribution of the measurement obtained from speech signals at low signal-to-noise ratios. It is shown that with the confidence measure, correct results can be effectively identified. By using the confidence measure, an adaptive algorithm for pitch estimation is proposed. Reliable pitch values obtained from preceding estimations are identified and used to predict the local pitch range for subsequent estimations. Parameter of the estimation algorithm is dynamically adjusted according to the predicted pitch range. Experimental results show that with the adaptive algorithm, pitch estimation accuracy is noticeably improved.

The performance of a speech recognition system degrades rapidly in the presence of ambient noise. To reduce the degradation, a degradation model is proposed which represents the spectral changes in a speech signal uttered in a noisy environment. The model uses frequency warping and amplitude scaling of each frequency band to simulate the variations of formant location, formant bandwidth, pitch, spectral tilt and energy in each frequency band by the Lombard effect. Another Lombard effect-the variation of overall vocal intensity-is represented by a multiplicative constant term depending on the spectral magnitude of the input speech. The noise contamination is represented by an additive term in the frequency domain. According to this degradation model, the cepstral vector of clean speech is estimated from that of noisy-Lombard speech using spectral subtraction, spectral magnitude normalization, band-pass filtering in the Lin-Log spectral domain, and multiple linear transformations. Noisy Lombard speech data is collected by simulating noisy environments using noises from automobiles, an exhibition hall, telephone booths in downtown crowded streets, and computer rooms. The proposed method significantly reduces error rates in the recognition of 50 Korean words. For example, the recognition rate is 95.91% with this method and 79.68% without this method at an SNR (signal-to-noise ratio) 10 dB.

In this paper, a method for the estimation of pitch from noise-corrupted speech observations based on extracting a pitch harmonic and the corresponding harmonic number is proposed. Starting from the harmonic representation of clean speech, a simple yet accurate harmonic sinusoidal autocorrelation (HSAC) model is first derived. By employing this HSAC model expressed in terms of the pitch harmonics of the clean speech, a new autocorrelation-domain least-squares fitting optimization technique is developed to extract a pitch harmonic from the noisy speech. Then, the harmonic number associated with the pitch harmonic is determined by maximizing an objective function formulated as an impulse-train weighted symmetric average magnitude sum function (SAMSF) of the noisy speech. The period of the impulse-train is governed by the estimated pitch harmonic and the maximization of the objective function is carried out through a time-domain matching of periodicity of the impulse-train with that of the SAMSF. An SAMSF-based pitch tracking scheme using dynamic programming is devised to obtain a smoothed pitch contour. In order to demonstrate the efficacy of the proposed method, simulations are conducted by considering naturally spoken speech signals in the presence of white or multi-talker babble noise at different signal-to-noise ratio (SNR) levels. A comprehensive evaluation of the pitch estimation results shows the superiority of the proposed method over some of the state-of-the-art methods under low levels of SNR.