Journal of the Acoustical Society of America最新文献

This paper presents a new turbo decision feedback equalizer and decoder (TDFED) for the orthogonal time-frequency space (OTFS) system of underwater mobile acoustic communications where the communication channel suffers from severe multipath and Doppler effects simultaneously. The proposed TDFED employs a set of feedforward and feedback filters in the time domain instead of the common approach that employs a normalized least mean square equalizer in the delay-Doppler domain. The receiver also utilizes low-complexity improved proportionate normalized least mean square channel estimation in the delay-Doppler domain. Practical OTFS modulation schemes are designed for acoustic transmission at a center frequency of 115 kHz and a symbol rate of 11.5 ksps (kilo-symbols-per-second). Several lake experiments in mobile communication scenarios are conducted to evaluate the proposed OTFS in comparison to the single-carrier coherent modulation (SCCM) and the orthogonal frequency division modulation (OFDM) schemes. The experimental results demonstrate that the proposed OTFS receiver effectively reduces the accuracy requirements of the Doppler compensation algorithm compared to the SCCM and OFDM schemes. The proposed TDFED algorithm achieves a much better bit error rate against long-multipath fading and severe Doppler shift than the existing delay-Doppler domain equalizers.

This article is focused on the viscous and inertial effects on airflow resistivity of periodic arrays of single-perforation plates spaced by thin air cavities. Analyzing this effect would provide better insight into losses within the material, including additional losses due to increasing sound excitation levels. In this way, the material pressure drop is predicted by computational fluid dynamics function (CFD) of the flow rate for corresponding pore Reynolds numbers between 0.3 and 1500. The static airflow resistivity coefficient is determined by the linear part of the pressure drop (viscous effect) and the Forchheimer coefficient from the nonlinear part of the pressure drop (inertial effect). Both coefficients are determined on the entirety of the material (globally) and at the plate levels (locally). Good agreement is observed between CFD predictions and experimental measurements on the whole range of studied Reynolds numbers. By locally investigating the pressure drops, the observations show that the viscous effects are constant through the material. With increasing pore Reynolds number, inertial effects of the first plate dominate over those of the other plates. The consideration of the local inertial effect will be a key component in the acoustic modeling of this type of material under high sound excitation levels.

Human speech recognition tends to be robust, despite substantial cross-talker variability. Believed to be critical to this ability are auditory normalization mechanisms whereby listeners adapt to individual differences in vocal tract physiology. This study investigates the computations involved in such normalization. Two 8-way alternative forced-choice experiments assessed L1 listeners' categorizations across the entire US English vowel space-both for unaltered and synthesized stimuli. Listeners' responses in these experiments were compared against the predictions of 20 influential normalization accounts that differ starkly in the inference and memory capacities they imply for speech perception. This includes variants of estimation-free transformations into psycho-acoustic spaces, intrinsic normalizations relative to concurrent acoustic properties, and extrinsic normalizations relative to talker-specific statistics. Listeners' responses were best explained by extrinsic normalization, suggesting that listeners learn and store distributional properties of talkers' speech. Specifically, computationally simple (single-parameter) extrinsic normalization best fit listeners' responses. This simple extrinsic normalization also clearly outperformed Lobanov normalization-a computationally more complex account that remains popular in research on phonetics and phonology, sociolinguistics, typology, and language acquisition.

Bilateral cochlear implant (BiCI) usage makes binaural benefits a possibility for implant users. Yet for BiCI users, limited access to interaural time difference (ITD) cues and reduced saliency of interaural level difference (ILD) cues restricts perceptual benefits of spatially separating a target from masker sounds. The present study explored whether magnifying ILD cues improves intelligibility of masked speech for BiCI listeners in a "symmetrical-masker" configuration, which ensures that neither ear benefits from a long-term positive target-to-masker ratio (TMR) due to naturally occurring ILD cues. ILD magnification estimates moment-to-moment ITDs in octave-wide frequency bands, and applies corresponding ILDs to the target-masker mixtures reaching the two ears at each specific time and frequency band. ILD magnification significantly improved intelligibility in two experiments: one with normal hearing (NH) listeners using vocoded stimuli and one with BiCI users. BiCI listeners showed no benefit of spatial separation between target and maskers with natural ILDs, even for the largest target-masker separation. Because ILD magnification relies on and manipulates only the mixed signals at each ear, the strategy never alters the monaural TMR in either ear at any time. Thus, the observed improvements to masked speech intelligibility come from binaural effects, likely from increased perceptual separation of the competing sources.

We investigate the impact of low-rank interference on the problem of distinguishing between two seabed types using ambient sound as an acoustic source. The resulting frequency-domain snapshots follow a zero-mean, circularly-symmetric Gaussian distribution, where each seabed type has a unique covariance matrix. Detecting changes in the seabed type across distinct spatial locations can be formulated as a two-sample hypothesis test for equality of covariance, for which Box's M-test is the classical solution. Interference sources such as passing ships result in additive noise with a low-rank covariance that can reduce the performance of hypothesis testing. We first present a method to construct a worst-case interference field, making hypothesis testing as difficult as possible. We then provide an alternating optimization procedure to recover the interference-free covariance matrix. Experiments on synthetic data show that the optimized interferer can greatly reduce hypothesis testing performance, while our recovery method perfectly eliminates this interference for a sufficiently small interference rank. On real data from the New England Shelf Break Acoustics experiment, we show that our approach successfully mitigates interference, allowing for accurate hypothesis testing and improving bottom loss estimation.

Three different automated procedures for efficiently measuring the 500 Hz N0Sπ binaural tone detection threshold of individuals with normal audiometric thresholds were tested with 6803 subjects and the results are compared with an automated version of an existing clinical procedure. Two of these procedures resulted in substantially reduced binaural detection performance and caused a notable decrease in the reliability of the behavior of the subjects. The remaining procedure was an 18-trial yes/no procedure and the difficulty of the trials in this procedure varied in a non-monotonic manner. This procedure not only has fewer trials than the existing clinical procedure but the reliability of the estimate of threshold is improved. The average difference in threshold between this procedure and the clinical procedure was only -0.67 dB, which is likely not clinically significant. Further, only 0.35% of the 6208 subjects tested with the non-monotonic-order procedure were unable to complete the fully automated test which was better than the failure rate for the automated version of the clinical procedure. With such a low failure rate, the modified procedure appears suitable for use as a rapid tool for helping to detect functional hearing deficits that are not captured by the audiogram.

The majority of existing piezoelectric transducers work at a single resonant frequency, and their applications in scenarios with multi-frequency or frequency variation are not fully considered. Moreover, emitting high-energy ultrasound at different frequencies is also crucial. Here, we propose the three-frequency coupled vibration piezoelectric transducer, which exhibits higher emission performances. The proposed transducer is comprised of two rectangular piezoelectric ceramics, which are cut from a piece of rectangular piezoelectric ceramic. We derive the three-dimensional coupled vibration electromechanical equivalent circuit of the proposed transducer. Then, the characteristics of the transducer are numerically simulated. And comparison experiments between the proposed transducer and a piece of rectangular piezoelectric ceramic transducer were done. An ultrasonic water tank measurement system was used to measure their sound field, axial sound pressure, and transmitting voltage response. Experiments are conducted to verify the electromechanical and sound field characteristics of transducers, which are in good agreement with the simulated results and theoretical predictions. The proposed transducer can generate stable and stronger energy ultrasonic waves at three resonant frequencies. And this study can provide the theoretical and experimental references for multi-frequency conversion and high-energy ultrasonic radiation of the transducer.

A robust multichannel speaker diarization and separation system is proposed by exploiting the spatiotemporal activity of the speakers. The system is realized in a hybrid architecture that combines the array signal processing units and the deep learning units. For speaker diarization, a spatial coherence matrix across time frames is computed based on the whitened Relative Transfer Functions of the microphone array. This serves as a robust feature for subsequent machine learning without the need for prior knowledge of the array configuration. A computationally efficient modified End-to-End Neural Diarization system in the Encoder-Decoder-based Attractor network is constructed to estimate the speaker activity from the spatial coherence matrix. For speaker separation, we propose the Global and Local Activity-driven Speaker Extraction network to separate speaker signals via speaker-specific global and local spatial activity functions. The local spatial activity functions depend on the coherence between the whitened Relative Transfer Functions of each time-frequency bin and the target speaker-dominant bins. The global spatial activity functions are computed from the global spatial coherence functions based on frequency-averaged local spatial activity functions. Experimental results have demonstrated superior speaker, diarization, counting, and separation performance achieved by the proposed system with low computational complexity compared to the pre-selected baselines.

This study presents a novel approach for estimating the transport parameters that characterize the acoustic behavior of fibrous materials using the Johnson-Champoux-Allard equivalent fluid model. We propose an inversion technique, based on an optimization algorithm, to fit the Johnson-Champoux-Allard model's predictions of normal incidence sound absorption coefficient to multi-compression-ratio experimental data. Experimental measurements using the two-microphone technique within an impedance tube are conducted on fibrous material samples tested at various compression ratios. Optimization is performed using both a non-linear programming solver and a genetic algorithm. Validation of the proposed method shows good agreement with well-established techniques and demonstrates its effectiveness across a range of fibrous materials. A sensitivity analysis emphasizes the importance of selecting appropriate boundaries for the search space in the optimization process. To enhance the robustness of optimization, a two-step iterative procedure is proposed. This straightforward methodology offers a robust and reliable framework for characterizing the transport properties of fibrous materials. Its ease of implementation and accuracy make it a valuable tool for enhancing material design and optimization in acoustic engineering.

Estimating the direction of arrival (DOA) under real-world conditions poses a significant challenge, as reverberations can lead to erroneous information. We note that the direct-path component and the reverberant components of sound exhibit distinct polarization states within the acoustic particle velocity field. Based on this observation, we propose and experimentally verify a method for localizing multiple sources in reverberant environments using a single acoustic vector sensor (AVS). The measurement of polarization states via AVS enables the identification of time-frequency bins primarily influenced by the direct-path component within the time-frequency domain, which are subsequently utilized for DOA estimation. Our study offers a novel perspective on sound field detection and may catalyze future applications including de-reverberation and the determination of environmental geometric parameters.