First-order Differential Microphone Research Articles

A speech separation algorithm based on the comb-filter effect

In the field of speech separation, the traditional single-channel and multi-channel speech separation methods have made great progress. However, the accuracy of separation and automatic speech recognition(ASR) rate are not yet satisfactory. With the development of neural networks, some scholars began to use deep learning to achieve speech separation. Although this kind of method improves the accuracy of speech separation, it also leads to the need for pre-training the model, higher computational complexity and reduced separation performance when the model does not match the mixed signal. This paper has conducted an in-depth study on the scene of multi-speaker separation, and proposed a new dual-channel speech separation algorithm based on the Comb-Filter Effect (CFE). The CFE is an effect that occurs when a signal passes through a first-order differential microphone(FDM) array. And this effect is discovered and exploited for the first time. By using this effect, this paper designed a new signal spectrum estimation method that can realize accurate estimation of speech signal, and combined this method with traditional spectral subtraction to achieve the purpose of speech separation. Finally, this paper compared the proposed algorithm with the traditional FastICA-based algorithm and the fully-convolutional time-domain audio separation network(Conv-TasNet)-based algorithm. The results of simulation and comparison experiments show that the algorithm can effectively separate two-way speech signals while greatly reducing the computational complexity and has excellent robustness. In various situations, the proposed algorithm can obtain the Scale-Invariant Source-to-Noise Ratio improvement (SI-SNRi) of 9.19 dB on average. In addition, the Short-Time Objective Intelligibility (STOI) and Perceptual Evaluation of Speech Quality (PESQ) of the speech signal can be improved by an average of 33% and 70% or more respectively.

Theoretical Lower Bounds on the Performance of the First-Order Differential Microphone Arrays With Sensor Imperfections

The first-order differential microphone arrays (DMAs) offer the appealing properties such as compact size, and frequency-invariant response, and thus have found a wide range of applications. However, they are known to be sensitive to sensor imperfections, i.e., microphone gain errors, phase errors and self-noise. In this paper, we study the analysis and design of the first-order DMAs in the presence of sensor imperfections. Particularly, the theoretical lower bounds on the performance of the first-order DMAs with sensor imperfections are derived in closed form by using the theory of interval analysis, including the lower bounds on directivity factor (DF) and front-to-back ratio (FBR), which are two essential performance measures in DMAs design. Unlike the existing related works, we have considered a more general and realistic scenario when all the above-mentioned three kinds of sensor imperfections coexist. Based on the derived performance lower bounds, a method is presented to design the first-order DMA with its performance exceeding the worst-case optimum performance. Moreover, an analysis is presented, which takes into account the sensor imperfections in the design of the first-order DMAs so the optimum worst-case performance is guaranteed. Extensive simulation results are also shown to demonstrate the effectiveness of our theoretical findings.

Speech endpoint detection in fixed differential beamforming combined with modulation domain

The actual speech signal is always in a complex environment, which affects the accuracy of speech endpoint detection. In order to improve the accuracy of speech endpoint detection under low Signal-To-Noise Ratio(SNR), a speech endpoint detection technology based on fixed differential beamforming and modulation domain is proposed. Firstly, two microphones of the first-order differential microphone arrays(DMAs) are used to collect the target speech and noise, and then a fixed differential beamformer is used to reduce the noise. Then, the phase compensated modulation domain spectral subtraction method is used to further eliminate the residual noise. Finally, the subband spectral entropy algorithm is used to detect the speech signal endpoint. The experimental results show that, compared with the existing methods, the proposed method has a higher accuracy of speech endpoint detection in the low SNR environment of -10∼5dB.

Deep learning assisted sound source localization using two orthogonal first-order differential microphone arrays.

Sound source localization in noisy and reverberant rooms using microphone arrays remains a challenging task, especially for small-sized arrays. Recent years have seen promising advances on deep learning assisted approaches by reformulating the sound localization problem as a classification one. A key to the deep learning-based approaches lies in extracting sound location features effectively in noisy and reverberant conditions. The popularly adopted features are based on the well-established generalized cross correlation phase transform (GCC-PHAT), which is known to be helpful in combating room reverberation. However, the GCC-PHAT features may not be applicable to small-sized arrays. This paper proposes a deep learning assisted sound localization method using a small-sized microphone array constructed by two orthogonal first-order differential microphone arrays. An improved feature extraction scheme based on sound intensity estimation is also proposed by decoupling the correlation between sound pressure and particle velocity components in the whitening weighting construction to enhance the robustness of the time-frequency bin-wise sound intensity features. Simulation and real-world experimental results show that the proposed deep learning assisted approach can achieve higher spatial resolution and is superior to its state-of-the-art counterparts using the GCC-PHAT or sound intensity features for small-sized arrays in noisy and reverberant environments.

A New Method to Design Steerable First-Order Differential Beamformers

First-order differential microphone arrays (FODMAs), which combine a small-spacing uniform linear array and a first-order differential beamformer, have been used in a wide range of applications for sound and speech signal acquisition. However, traditional FODMAs are not steerable and their main lobe can only be at the endfire directions. To circumvent this problem, we propose in this letter a new method to design steerable FODMAs. We first divide the target beampattern into a sum of two sub-beampatterns, i.e., cardioid and dipole, where the summation is controlled by the steering angle. We then design two sub-beamformers, one is similar to the traditional approach and is used to achieve the cardioid sub-beampattern, while the other is designed to filter the squared observation signals and is used to approximate the dipole sub-beampattern. The overall beampattern resembles the target beampattern for any steering angle. Simulations and experiments are performed to justify the effectiveness of the developed method.

Wave Digital Implementation of Robust First-Order Differential Microphone Arrays

In this letter, a novel time-domain implementation of robust first-order differential microphone arrays (DMAs), based on wave digital filters, is presented. The proposed beamforming method is extremely efficient, as it requires at most two multipliers and one delay for each filter, where the necessary number of filters equals the number of physical microphones of the array, and it avoids the use of fractional delays. The update of the coefficients of the filters, required for reshaping the beampattern, has a significantly lower computational cost with respect to the time-domain methods presented in the literature. This makes the proposed method suitable for real-time DMA applications with time-varying beampatterns.

DOA estimation of multiple speech sources by selecting reliable local sound intensity estimates

Sound source DOA estimation using first-order differential microphone arrays (DMAs) has been demonstrated as a promising means for the applications where the size of arrays is restricted. The existing methods for DOA estimation of multiple speech sources with first-order DMAs, however, are shown sensitive to noise and room reverberation. To combat the problem, we propose a DOA estimation algorithm by exploring the redundancies of two orthogonal first-order DMAs in sound intensity measurement. In particular, the reliable time–frequency points for DOA estimation can be effectively singled out by the proposed algorithm and thus leads to better DOA estimation performance in noisy and reverberant environments. Moreover, the proposed algorithm has a closed form solution, and no time-consuming search process over spatial space is required. Simulation and real experimental results have demonstrated the effectiveness of the proposed algorithm.

Noise-Reducing and Effect-Enhancing System for Communication Equipment Mounted on Vehicle

The noise of communication equipment mounted on Vehicle is one of important factors affecting equipment efficiency. Active Noise Control (ANC) technology is applied to noise reduction of the equipment, and a noise reduction mathematics model is built based on accurate noise measurement. An approach to reduce noise based on first-order differential microphone (FDM) array and ANF technology is proposed, which collects and evaluates environmental noise precisely to enhance voice. This approach improves the conversation effects and realizes full fidelity of voice transmission.

KEMAR 마네킹을 이용한 단이 보청기용 FDSI 빔포밍 알고리즘의 정량적 평가

Abstract: To enhance the speech perception of hearing aid users in noisy environment, most hearing aid devicesadopt various beamforming algorithms such as the first-order differential microphone (DM1) and the two-stage direc-tional microphone (DM2) algorithms that maintain sounds from the direction of the interlocutor and reduce the ambi-ent sounds from the other directions. However, these conventional algorithms represent poor directionality abilityin low frequency area. Therefore, to enhance the speech perception of hearing aid uses in low frequency range, ourgroup had suggested a fractional delay subtraction and integration (FDSI) algorithm and estimated its theoretical per-formance using computer simulation in previous article. In this study, we performed a KEMAR test in non-rever-berant room that compares the performance of DM1, DM2, broadband beamforming (BBF), and proposed FDSIalgorithms using several objective indices such as a signal-to-noise ratio (SNR) improvement, a segmental SNR (seg-SNR) improvement, a perceptual evaluation of speech quality (PESQ), and an Itakura-Saito measure (IS). Experi-mental results showed that the performance of the FDSI algorithm was −3.26-7.16 dB in SNR improvement, −1.94-5.41 dB in segSNR improvement, 1.49-2.79 in PESQ, and 0.79-3.59 in IS, which demonstrated that the FDSI algo-rithm showed the highest improvement of SNR and segSNR, and the lowest IS. We believe that the proposed FDSIalgorithm has a potential as a beamformer for digital hearing aid devices.Key words: hearing aids, directional microphone, beamforming

Aspects of first‐order differential microphone arrays in the presence of sensor imperfections

AbstractArray processing for broadband signals like speech is not a trivial task. Most sensor arrangements and the related beamforming algorithms show strong frequency dependent characteristics. Differential microphone arrays promise frequency independent high directional gain with compact arrangements. However, this type of microphone array has also the problem that even small deviations in microphone properties can cause severe degradation of the array's performance. Especially at low frequencies the achievable gain is limited by the influence of sensor mismatch and sensor noise.In this contribution the performance of first‐order differential arrays is examined and important parameters such as array response and directivity index are thoroughly studied. The frequency dependence of these parameters is investigated in order to derive limits for a favorable frequency range. First, the performance of the array is examined for ideal sensors. Then, two different models for sensor mismatch are introduced to describe the degradation of the array's performance. In addition, solutions for sensor calibration and experimental results with real microphones are presented.

Image-derived second-order differential microphones

Second-order differential toroidal and unidirectional microphones derived using a bipolar first-order differential sensor and a reflecting plane are described. The bipolar first-order differential microphone is positioned with its axis perpendicular to and suspended a few centimeters from a large acoustically reflecting surface. The resulting sensor image is phase reversed resulting in a transducer that is a linear quadrupole. The linear quadrupole can be described by two dimensions, the distance corresponding to the dipole distance and twice the distance from the reflecting plane. If the reflecting surface is large enough or if the wall of a room is used, the resulting microphone becomes a second-order differential unidirectional microphone. The match between the sensor and its image from a good acoustic reflector results in a second-order differential microphone with 3-dB beamwidth of ±33° and no grating lobes below 3 kHz for a 2.5-cm spacing between the sensor and reflecting surface. A wall-mounted toroid can be formed by using two bipolar first-order sensors at right angles to each other and with the axis of each sensor at 45° to the reflecting surface and a spacing between transducers that is twice the height of the transducers from the reflecting plane. A table-mounted toroid can be realized by properly combining a filtered version of a suspended first-order sensor and an omnidirectional sensor flush mounted to the reflecting table top.