Adaptive Noise Cancellation to Speech Enhancement

Adaptive dual-sensor noise cancellation method for continuous wave mud pulse telemetry

MRS (Magnetic Resonance Sounding)) method is a direct method for underground water detecting which is easily disturbed by interferences according to its weak signal (nV). Time domain quantization methods require a good suppression of interferences for more accurate MRS signal parameter estimation. Of all interferences, power line harmonic interference is the most significant. To solve the problem, the paper analyzes the features of the MRS signal and the interference types and discusses two easily program integrated interference elimination method for MRS signal real time pretreatment. The stacking method based on random phase is an effective way to eliminate the interference but it requires massive stacks. Another method, Adaptive Noise Cancellation (ANC) method is also proposed which is a more effective method especially in power line harmonic interference cancellation. Here an improved bi- direction ANC method is proposed. The method can significantly eliminate the power line harmonic interference. Field test was carried out using JLMRS-I developed by Jilin University. With the method, the SNR of free induction decay (FID) records can be improved by 4 times with stacks 20 and nearly 100 times with stacks 160. The application of the bi-direction ANC method can improve the exploration efficiency under the same SNR of the FID records.

Line noise is artifact affecting a gamma band of the EEG. Conventional filters are used for removing the line noise artifact, but these filters, also remove physiological activity and could induce spurious oscillations. We tested two alternative methods for removing line noise artifact without removing a physiological activity, specifically Adaptive noise cancellation (ANC) method with Linear regression and modified Independent Component Analysis (ICA). We proposed a new protocol for statistical evaluation of line noise removal effectiveness. The line noise artifact was physically simulated and the protocol for statistical evaluation was verified. The ANC method shows good results according to statistical evaluation. The lowest residuum of simulated line noise artifact corresponds to 0.0005 \(\upmu \)V\(^2\)/Hz. Nevertheless, the ANC is very sensitive to the quality of the reference signal, which heavily depends on the reference electrode selection.

Speech enhancement is performed in a wide and varied range of instruments and systems. In this paper, a novel approach to Speech Enhancement using Adaptive Empirical Mode Decomposition (SEAEMD) is presented. Spectral analysis of non-stationary signals can be performed by employing techniques such as the STFT and the Wavelet transform (WT), which use predefined basis functions. Empirical Mode Decomposition (EMD) performs very well in such environments. EMD decomposes a signal into a finite number of data-adaptive basis functions, called Intrinsic Mode Functions (IMFs). The new SEAEMD system incorporates this multi-resolution approach with adaptive noise cancellation (ANC) for effective speech enhancement on an IMF level, in stationary and non-stationary noise environments. A comparative performance study is included that compares the competitive method of conventional ANC to the robust SEAEMD system. The results demonstrate that the new system achieves significantly improved speech quality with a lower level of residual noise.

Pulse rate estimation from forehead photoplethysmograph signal using RLS adaptive filtering with dynamical reference signal

Objective: Aiming at the vibration control problem existing in the vector hydrophone under platform vibration and noise excitation, two methods, adaptive noise cancellation algorithm and vibration reduction design, are used to suppress vibration and noise. Methods: A simulation model of vibration and noise signals in the side area of the platform is established first, and the simulation model is simulated by adaptive noise cancellation method. Then noise reduction simulation is carried out by seismic mitigation and isolation method. Finally, the simulation results of the two methods are compared to check the effectiveness of the two noise suppression methods. Results: The results show that adaptive noise cancellation algorithm can effectively cancel the line spectra of platform vibration and noise, and the reduction of continuous spectra is about 10 dB; while the vibration reduction method can reduce the whole vibration and noise of the model, and the vibration and noise interference of the vector hydrophone is reduced by about 15 dB. Conclusion: The obtained results can serve as valuable reference for the performance optimization of suppression methos for vector hydrophones under platform’s vibration and noise excitation.

Dual-microphone active noise cancellation paved with Doppler assimilation for TADS

In this paper, we propose a novel speech enhancement (SE) method by exploiting the discrete wavelet transform (DWT). This new method reduces the amount of fast time-varying portion, viz. the DWT-wise detail component, in the spectrogram of speech signals so as to highlight the speech-dominant component and achieves better speech quality. A particularity of this new method is that it is completely unsupervised and requires no prior information about the clean speech and noise in the processed utterance. The presented DWT-based SE method with various scaling factors for the detail part is evaluated with a subset of Aurora-2 database, and the PESQ metric is used to indicate the quality of processed speech signals. The preliminary results show that the processed speech signals reveal a higher PESQ score in comparison with the original counterparts. Furthermore, we show that this method can still enhance the signal by totally discarding the detail part (setting the respective scaling factor to zero), revealing that the spectrogram can be down-sampled and thus compressed without the cost of lowered quality. In addition, we integrate this new method with conventional speech enhancement algorithms, including spectral subtraction, Wiener filtering, and spectral MMSE estimation, and show that the resulting integration behaves better than the respective component method. As a result, this new method is quite effective in improving the speech quality and well additive to the other SE methods.

In recent years, speech enhancement in the presence of very harsh noise has received a lot of attention. In this paper, an adaptive noise canceller (ANC) having a single-reference and multi-primary channels (SRMP) is introduced to recover the speech signal from its contaminated measurements. This SRMPANC takes a bone-conducted (BC) signal as its reference and adopts multiple noisy air-conducted (AC) speech measurements as its primary signals. The multiple primary channels are connected by a linear combiner (LC) which is optimized and updated by a recursive least square (RLS) algorithm. A linear or a nonlinear filter (e.g. FIR, Volterra, etc.) is equipped in the SRMP-ANC, that is updated by an LMS or LMS-like algorithm. Extensive simulations with real bone- and air-conducted speech measurements are performed to demonstrate the effectiveness of the proposed SRMP-ANC. It is revealed that the speech recovery quality is significantly improved as the number of primary channels is increased in terms of both mutual correlation and mean square error (MSE) between the AC speech signal and its recovered version.

This paper proposes novel heuristic algorithm called gravitational search algorithm (GSA) to speech enhancement. Stochastic and heuristic algorithms like particle swarm optimization (PSO) and some of its variants have been adapted to the field of speech enhancement in recent years. Although standard PSO (SPSO) finds good solutions, it suffers from the premature convergence by getting trapped into local optimum. In order to increase the diversity in search space and to improve the local searching capability, another recently developed heuristic algorithm GSA is proposed to speech enhancement in this paper. GSA is mainly constructed on the basis of law of gravity and the notion of mass interactions .The proposed algorithm is studied for real world noise condition called babble noise, at three different input SNR levels. To the best of our knowledge, there is no analysis about the intelligibility of enhanced speech using optimization techniques. In the present study, the proposed algorithm is compared with the standard PSO (SPSO) algorithm for dual channel speech enhancement, and the intelligibility analysis is also reported. Simulation results indicate that GSA-based algorithm outperforms the particle swarm optimization in adaptive noise cancellation with an improved speech quality and intelligibility.

Most algorithms for speech enhancement in the spectral domain focus on the acquisition of an estimator of the clean speech parameter, such as spectrum or amplitude. Enhanced speech quality with these methods mainly depends upon the accuracy of the estimator. When the signal-to-noise ratio (SNR) becomes lower, e.g., less than 3 dB, the enhanced speech often shows unsatisfactory quality. Here, we propose a new method for speech enhancement in the spectral domain. It is based on our view that speech characteristics are perceived mainly by part of the spectral components which have higher local (or instantaneous) SNRs. These components have a special importance on speech SNR enhancement and speech understanding. Identifying these components in the noisy spectrum with a decision system and constructing enhanced speech just by the same noisy amplitude and phase, the result is similar to those popular algorithms. All that is needed is a local SNR decision for each spectral component. The threshold for the decision is derived and two popular methods for local SNR (LSNR) estimation are suggested. It shows better performance under lower SNR signals. Though the number of these kind of spectral components varies with different signal SNR, any algorithms which can pick up more than 90% of it is sufficient to match any existing methods. Primary considerations and results are shown.

The generalized sidelobe cancellation (GSC) algorithm is a widely used microphone array speech enhancement method. Compared with the conventional single-channel speech enhancement, microphone array algorithm can effectively utilize the spatial information of the signal to better suppress the noise. The conventional GSC algorithm is prone to cause the desired signal to be cancelled when adaptively removing noise, and also has low robustness when the adaptive filter is updated. This paper presents a new adaptive filter coefficient updating algorithm for GSC. First, the output of the superdirective beamformer (SDB) is used to find the coherence with the original signal, and then the output of the delay-sum beamformer (DSB) and blocking matrix (BM) is used to find the signal-to-interference ratio (SIR) in a certain frequency range. Finally, the obtained coherence and SIR are combined to control the updating of the adaptive noise canceller (ANC). The experimental results show that the proposed algorithm not only further suppresses different noise interferences at different directions, but also has stronger robustness than the conventional GSC algorithm.

Speech enhancement is a process which cultivates the quality of speech signal in noisy environment. It refers to removing or reducing the background noise in order to obtain an improved quality of original speech signal. Degradation of speech signal is most common problem in speech communication, so enhancement of the speech plays a vital role in improving the quality of speech signal. A number of methods are used for speech enhancement. Here we are using LPC based FEM in Wiener filter for speech enhancement. This method is then compared with several speech enhancement algorithms to obtain a better result or better speech quality. On comparing this method with other speech enhancement methods we obtain better speech performance. Here we are using NOIZEUS speech database in order to compare different speech enhancement methods. The experimental result shows that our proposed method provides better result and there is no information loss in original speech signal.

Speech enhancement (SE) methods mainly focus on recovering clean speech from noisy input. In real-world speech communication, however, noises often exist in not only speaker but also listener environments. Although SE methods can suppress the noise contained in the speaker’s voice, they cannot deal with the noise that is physically present in the listener side. To address such a complicated but common scenario, we investigate a deep learning-based joint framework integrating noise reduction (NR) with listening enhancement (LE), in which the NR module first suppresses noise and the LE module then modifies the denoised speech, i.e., the output of the NR module, to further improve speech intelligibility. The enhanced speech can thus be less noisy and more intelligible for listeners. Experimental results show that our proposed method achieves promising results and significantly outperforms the disjoint processing methods in terms of various speech evaluation metrics.

Deep learning-based speech enhancement (SE) methods predominantly draw upon two architectural frameworks: generative adversarial networks and diffusion models. In the realm of SE, capturing the local and global relations between signal frames is crucial for the success of these methods. These frameworks typically employ a UNet architecture as their foundational backbone, integrating Long Short-Term Memory (LSTM) networks or attention mechanisms within the UNet to effectively model both local and global signal relations. However, the coupled relation modeling way may not fully harness the potential of these relations. In this paper, we propose an innovative Association-based Fusion Speech Enhancement method (AFSE), a decoupled method. AFSE first constructs a graph that encapsulates the association between each time window of the speech signal, and then models the global relations between frames by fusing the features of these time windows in a manner akin to graph neural networks. Furthermore, AFSE leverages a UNet with dilated convolutions to model the local relations, enabling the network to maintain a high-resolution representation while benefiting from a wider receptive field. Experimental results demonstrate that the AFSE method significantly improves performance in speech enhancement tasks, validating the effectiveness and superiority of our approach. The code is available at https://github.com/jie019/AFSE_IJCAI2025.