Archives of Acoustics

Fidgety speech emotion has important research value, and many deep learning models have played a good role in feature modeling in recent years. In this paper, the problem of practical speech emotion is studied, and the improvement is made on fidgety-type emotion using a novel neural network model. First, we construct a large number of phonological features for modeling emotions. Second, the differences in fidgety speech between various groups of people were studied. Through the distribution of features, the individual features of fidgety emotion were studied. Third, we propose a fine-grained emotion classification method, which analyzes the subtle differences between emotional categories through Siamese neural networks. We propose to use multi-scale residual blocks within the network architecture, and alleviate the vanishing gradient problem. This allows the network to learn more meaningful representations of fidgety speech signal. Finally, the experimental results show that the proposed method can provide the versatility of modeling, and that fidgety emotion is well identified. It has great research value in practical applications.

Mismatch negativity (MMN) essentially reflects auditory change detection. Although auditory change detection can potentially be assessed through behavioral auditory testing methods, the increased reliability of objective methods, such as MMN, makes them more valuable. The aim of this study was to detect and compare the intensity just noticeable difference using the MMN and a behavioral method. The level at which the intensity difference between the frequent stimulus and the infrequent stimulus was the lowest and the MMN wave elicited was accepted as the MMN threshold. A total of 60 subjects, 30 females (mean age 21.70, SD = 1.91 years) and 30 males (mean age 22.77, SD = 3.01), aged 20–30 years, were included in the study. In the whole sample, a significant difference was found between MMN thresholds obtained from the right ear side and MMN thresholds obtained from the left ear side, regardless of sex (p < 0:05). In the comparison of the values obtained using the behavioral method and MMN, no significant difference was found for either the right or the left side in both sexes (p > 0:05). The results showed that the values determined by the behavioral method and MMN on both the right and left ear sides were similar in both sexes.

The signature of fifths is a special kind of music representation technique enabling acquisition of musical knowledge. The technique is based on geometrical relationships existing between twelve polar vectors inscribed in the circle of fifths, which represent individual pitch-classes detected in a given composition. In this paper we introduce a real-time key-detection algorithm which utilizes the concept of the signature of fifths. We explain how to create the signature of fifths and how to derive its descriptors required by the algorithm, i.e., the main directed axis of the signature of fifths, the major/minor mode axis, the characteristic vector of the signature of fifths, the characteristic angle of the signature of fifths, and the angle of the major/minor mode. We performed a series of experiments to test the algorithm’s effectiveness. The results were compared with those obtained using key-detection approaches based on key-profiles. All experiments were conducted using works composed by J.S. Bach, F. Chopin, and D. Shostakovich. The distinctive features of the presented algorithm, with respect to the considered key-detection approaches, are computational simplicity and stability of the decision-making process.

Vehicle engine vibration signals acquired using MEMS sensors are crucial in the diagnosis of engine malfunctions, notably misfires due to unwanted signals and external noises in the recorded vibration dataset. In this study, the ADXL1002 accelerometer interfaced with the Beaglebone Black microcontroller is employed to capture vibration signals emitted by the vehicle engine across various operational states, including unloaded, loaded, and misfire conditions at 1500 RPMs, 2500 RPMs, and 3000 RPMs. In conjunction with the acquisition of this raw vibration data, frequency-domain signal processing techniques are employed to meticulously analyze and diagnose the distinct signatures of misfire occurrences across various engine speeds and loads. These techniques encompass the fast Fourier transform (FFT), envelope spectrum (ES), and empirical mode decomposition (EMD), each tailored to discern and characterize the nuanced vibration patterns associated with misfire events at different operational conditions.

This paper presents the results of preliminary research aimed at developing a method for rapid, noncontact diagnostics of the electric drive of car seats. The method is based on the analysis of acoustic signals produced during the operation of the drive. Pattern recognition and machine learning processes were used in the diagnosis. A method of feature extraction (diagnostic symptoms) using wavelet decomposition of acoustic signals was developed. The discriminative properties of a set of diagnostic symptoms were tested using the “Classification Learner” application available in MATLAB. The obtained results confirmed the usefulness of the developed method for the technical diagnostics of car seats.

redict the sound quality of dual-phase Hy-Vo chain transmission system noise using a small sample size. Noise acquisition tests are conducted under various working conditions, followed by subjective evaluations using the equal interval direct one-dimensional method. Objective evaluations are performed using the Mel-frequency cepstral coefficient (MFCC). To understand the impact of the MFCC order and the frame number on prediction accuracy, MFCC feature maps of different specifications are analyzed. The dataset is expanded threefold using fuzzy generation with an appropriate membership degree. The convolutional neural network (CNN) is developed, utilizing MFCC feature maps as inputs and evaluation scores as outputs. Results indicate a positive correlation between the frame number and prediction accuracy, whereas higher MFCC orders introduce redundancy, reducing accuracy. The proposed CNN method outperforms three traditional machine learning approaches, demonstrating superior accuracy and resistance to overfitting.

This study evaluated the combined sensitivity analysis of several room acoustic descriptors: reverberation time (T30), center time (Ts), early decay time (EDT), definition (D50), clarity (C50), useful-to-detrimental sound ratio (U50), and speech transmission index (STI); and also it assessed how these descriptors responded jointly to different acoustic-structural factors. The first-order factors were background noise (A), acoustic ceiling tile sound absorption coefficient (B), confinement (C), and occupancy (D), along with their interaction effects. A novel method is proposed for this joint evaluation of sensitivity factors. This method involves in situ measurements and an unreplicated 24 factorial design, which has been validated by ODEON software. The significance of input factors is determined using artificial neural networks (ANN) and the modified profile method (MPM), validated by multiple linear regression (MLR). Three significant correlation groups are identified at p < 0:05: group 1 (EDT, T30, Ts), group 2 (C50, D50), and group 3 (U50, STI). The ceiling material sound absorption (B) is found to affect reverberation (groups 1 and 2), while background noise (A) impacts STI and U50. A weak correlation is found between D50 and STI. These results are confirmed by the MLR and MPM methods.

It has been shown that within the range of acoustic pressures used in ultrasound imaging, waveforms are distorted during propagation in tissue due to the physically nonlinear behavior of the tissue. This distortion leads to changes in the spectrum of the received ultrasound echoes, causing the transfer of signal energy from the fundamental frequency to higher harmonics. Interestingly, adipose tissue exhibits up to 50 % stronger nonlinear behavior compared to other soft tissues. The tissue nonlinearity parameter B/A is typically measured ex vivo using an ultrasound method in transmission mode, which requires extensive receiving systems. Currently, there is no improved ultrasound method for measuring the B/A nonlinearity parameter in vivo, which could be used in assessing the degree of fatty liver disease. We propose a new, simple approach to estimating nonlinear tissue properties. The proposed method involves transmitting ultrasound waves at significantly different acoustic pressures, recording echoes only in the fundamental frequency band at various depths, and introducing a nonlinearity index (NLI) based on specific echo amplitude ratios. The NLI at a given depth is calculated using the ratio of two dimensionless parameters. The first parameter is a predetermined constant obtained by dividing the total echo values from transmitting a signal at higher sound pressure by those from a signal at lower sound pressure, summed over a small tissue sample volume located near the transducer. The second parameter is calculated at a fixed distance from the transducer, determined by dividing the total echo values from transmitting a signal at higher sound pressure by those from a signal at lower pressure, summed over a small tissue volume of the tissue at that distance from the transducer. The reliability of the proposed measurements for assessing tissue nonlinearity has been substantiated through experimental confirmation of the existing correlations between the values of NLI and B/A in water, sunflower oil, and animal liver tissue samples with oil-enriched regions. The NLI was more than 15 % higher in sunflower oil than in water. The NLI in bovine liver sample below the area with injected oil (mimicking “steatosis”) was more than 35 % higher than in regions without oil. This method represents a promising modality for the nonlinear characterization of tissue regions in vivo, particularly for diagnosing fatty liver disease.