Sound Signals Research Articles

A method of sound source recognition based on two stagewise weak orthogonal matching pursuit

The stagewise weak orthogonal matching pursuit (SWOMP) algorithm solves the problem that the orthogonal matching pursuit (OMP) algorithm relies too much on the sparsity of sound sources in the process of sound source identification. However, compared with OMP algorithm, SWOMP algorithm has some disadvantages, such as lower spatial resolution and worse identification accuracy. To solve the problem of poor reconstruction effect of SWOMP algorithm, a sound source identification algorithm based on two stagewise weak orthogonal matching pursuit (TSWOMP) is proposed. In this method, the initial atom set is obtained by threshold selection, the reliability of the primary atom set is detected, and the wrong atoms in the primary atom set are further deleted. Finally, the position and size of the sound source signal are identified according to the obtained final atom set. Under different frequencies and signal-to-noise ratios, the simulation results show that compared with SWOMP algorithm, TSWOMP algorithm can effectively reduce the main lobe width, suppress the influence of side lobes and significantly improve the spatial resolution. Moreover, even if the sparsity of sound source is unknown, it can achieve the same recognition accuracy as OMP algorithm and has high recognition stability. Finally, the validity and reliability of TSWOMP algorithm are further verified by experiments.

Improving the Diagnosis of Cardiac Abnormalities Through Feature Extraction from the Heart Sound Signal Using Machine Learning Classification Algorithms

Improving the Diagnosis of Cardiac Abnormalities Through Feature Extraction from the Heart Sound Signal Using Machine Learning Classification Algorithms

Verifying turbulence models in the atmospheric boundary layer by acoustical means

The impacts of turbulence on sound propagation in the atmospheric boundary layer (ABL) are important in many practical applications such as acoustic source localization, sonic boom propagation, and auralization of flying aircraft. However, turbulence impacts can vary dramatically in response to changing meteorological conditions. To address this issue, a turbulence spectral model has been introduced that captures the complexities of turbulence generation by wind shear, buoyancy instabilities, and ground blocking in the ABL. An experiment on vertical and slanted sound propagation through the ABL was conducted to verify the model. Various statistical characteristics of sound signals, including variances of the log-amplitude and phase fluctuations, coherences, and signal probability distributions were measured and compared to theoretical predictions. The dependence of the statistics on meteorological conditions was found to be accurately predicted in conditions of well-developed turbulence, as occurs in windy conditions and the daytime. Predictions in conditions of weak and intermittent turbulence, as often occur at night and around sunrise and sunset, remain challenging.

Theoretical model for correlation analysis of wideband signal received by two vertical arrays and retrieval of modal dispersion curves in shallow water

Goal of this work is to carry out theoretical modeling extraction of characteristics of layered bottom (sound speeds c ;2 in layer, c in water, c ;b in half-space and layer’s thickness h using construction of correlation function (matrix) of wideband synthetic sound signal radiated by the moving source and imitating shipping noise. This work's motivation was experiment carried out by authors in shallow water. Combination of two vertical line arrays acts as a diffraction grating with 2N hydrophones that uses interference pattern in vertical and horizontal directions and gives more information than one-dimensional array or single receiver. Our waveguide model constitutes water layer above low-speed thin layer (c ;2 ≪ c, h is about a few meters) and half space (c ;b > c). In this waveguide there are two sorts of modes: in dependence on frequency and thickness of sediment’s layer– modes with vertical dependence distributed in water layer and modes trapped in the near bottom layer. It is shown that correlation function constructed in coordinates (phase speed, frequency) allows us to retrieve dispersion curves and to estimate critical frequencies—determining creation of trapped modes, which in turn give parameters of layered bottom. [Work was supported by ISF, Grant No. 946/20.]

The Average Sound Energy Spectrum Produced by a Single Raindrop

The characteristics of underwater acoustic radiation generated by a single raindrop impacting the water surface play a fundamental role in the study of rainfall noise. However, the underwater sound signal generated by a single raindrop has great randomness in the amplitude of sound pressure, frequency range and energy size, so that the uncertainty of the sound energy spectrum curve of a single raindrop is relatively large. In this paper, by generating artificial rainfall containing only one size of raindrop, the acoustic radiation generated by artificial rainfall composed of about 136,000 raindrops with an equivalent diameter of 0.9 mm was measured by reverberation method. A smooth, continuous sound energy spectrum of single raindrop covering the full frequency band is obtained.

Effect of underwater vehicle wake on sound propagation characteristics in stratified medium

Due to the uneven distribution of seawater temperature and salinity, the density distribution at different depths is different, thus forming a density step layer with a certain gradient. Acquiring the wake internal wave field generated by underwater vehicles passing through this layered structure and mastering its modulation characteristics for the sound field can provide new technical support for the tracking and identification of underwater vehicles. In this paper, the real size model of a certain type of underwater vehicle is taken as the research object to explore the characteristics of its wake internal wave field, and the influence of navigation parameters on the wake field and the change of winning characteristics under the influence of internal waves are simulated and analyzed. The results show that the navigation parameters of underwater vehicles have obvious influence on the wake field, and then have specific influence on the sound field structure in a certain range, especially on the sound signal of the sound field.

Diagnosis of pulmonary hypertension associated with congenital heart disease based on statistical features of the second heart sound

Aiming at the problems of obscure clinical auscultation features of pulmonary hypertension associated with congenital heart disease and the complexity of existing machine-aided diagnostic algorithms, an algorithm based on the statistical characteristics of the high-frequency components of the second heart sound signal is proposed. Firstly, an endpoint detection adaptive segmentation method is employed to extract the second heart sounds. Subsequently, the high-frequency component of the heart sound is decomposed using the discrete wavelet transform. Statistical features including the Hurst exponent, Lempel-Ziv information and sample entropy are extracted from this component. Finally, the extracted features are utilized to train an extreme gradient boosting algorithm (XGBoost) classifier, which achieves an accuracy of 80.45% in triple classification. Notably, this method eliminates the need for a noise reduction algorithm, allows for swift feature extraction, and achieves effective multi-classification using only three features. It is promising for early screening of pulmonary hypertension associated with congenital heart disease.

Microwave Speech Recognizer Empowered by a Programmable Metasurface.

Speech recognition becomes increasingly important in the modern society, especially for human-machine interactions, but its deployment is still severely thwarted by the struggle of machines to recognize voiced commands in challenging real-life settings: oftentimes, ambient noise drowns the acoustic sound signals, and walls, face masks or other obstacles hide the mouth motion from optical sensors. To address these formidable challenges, an experimental prototype of a microwave speech recognizer empowered by programmable metasurface is presented here that can remotely recognize human voice commands and speaker identities even in noisy environments and if the speaker's mouth is hidden behind a wall or face mask. The programmable metasurface is the pivotal hardware ingredient of the system because its large aperture and huge number of degrees of freedom allows the system to perform a complex sequence of sensing tasks, orchestrated by artificial-intelligence tools. Relying solely on microwave data, the system avoids visual privacy infringements. The developed microwave speech recognizer can enable privacy-respecting voice-commanded human-machine interactions is experimentally demonstrated in many important but to-date inaccessible application scenarios. The presented strategy will unlock new possibilities and have expectations for future smart homes, ambient-assisted health monitoring, as well as intelligent surveillance and security.

IOT Rancang Bangun Alat Pengusir Hama Burung pada Padi Sawah Petani Berbasis Internet of Things (IoT)

This research aims to develop an effective and efficient bird pest repellent tool in farmers' rice fields by utilizing Internet of Things (IoT) technology. Bird pests often pose a serious threat to agricultural products, especially in rice fields. Through the application of IoT, this tool can monitor real-time bird activity around agricultural areas and take appropriate preventive measures. This tool is designed to utilize motion sensors or pir sensors, surveillance cameras, and a sound or visual signal delivery system to chase away birds. approaching agricultural areas. PIR sensors are widely used in a variety of applications, including home security systems, motion sensor lighting, and building automation. The main advantages of PIR sensors are their low power consumption and ability to provide fast response to movement, making them an efficient and effective choice for many detection applications. motion, data collected from sensors and cameras will be sent to the IoT platform with the Telegram application and which is connected via an internet connection, allowing farmers to monitor and control tools remotely via smart devices such as smartphones or computers. The main advantage of this tool is its ability to provide a quick response to the presence of bird pests, reduce agricultural yield losses and increase crop productivity. Additionally, integration with IoT provides high flexibility and connectivity, enabling continuous system optimization. designing this tool involves the hardware and software design stages. The hardware consists of a pir sensor, an esp 32 cam microcontroller, and a speaker and servo. While the software consists of Arduino Ide, Blynk and Telegram. After design, the tool is field tested for performance. Test results show that this tool is able to detect the presence of birds with good accuracy, provides a fast response in chasing them away, Telegram is able to communicate by giving commands and receiving images, Blynk can provide information on the condition of the tool when it is online and give commands to activate the servo, sound or both of them. With a significant removal success rate, this tool is successful in reducing losses caused by bird strikes in agriculture. Additionally, it has low power consumption and can be integrated with existing IoT platforms for remote monitoring and data analysis. It is hoped that this research can make a positive contribution to efforts to develop sustainable agricultural technology and improve farmers' welfare by protecting agricultural products from bird pest attacks. Thus, it is hoped that the implementation of this tool in the field can provide an innovative and efficient solution in controlling bird pests in farmers' rice fields.

Acoustic vibration multi-domain images vision transformer (AVMDI-ViT) to the detection of moldy apple core: Using a novel device based on micro-LDV and resonance speaker

Moldy-core is a common internal disease in apples, and apples infected with this disease cannot be directly identified according to their external characteristics. In this study, a novel acoustic vibration device based on micro-LDV, resonance speaker and microphone was employed to detect moldy-core in apples, the acoustic vibration signals of healthy apples and apples with different degrees of moldy-core were converted into acoustic vibration multi-domain images (AVMDI), which consisted of time-frequency images generated through continuous wavelet transform (CWT), as well as time-domain and frequency-domain images generated through Gramian Angular Field (GAF). Subsequently the combination of AVMDI and Vision Transformer (ViT) was applied to the identification of internal defects in fruits. The outcomes evince that the classification efficacy of the model, amalgamating sound and vibration signals, surpasses that of models reliant on solitary sound or vibration signals. The AVMDI-ViT model achieved an overall classification accuracy of 97.96 %. Specifically, it achieved 100 % accuracy in identifying healthy apples, 94.74 % accuracy in identifying mild moldy-core apples (≤ 7 %), 97.50 % accuracy in identifying moderate moldy-core apples (> 7 % and ≤ 15 %), and 100 % accuracy in identifying severe moldy-core apples (> 15 %). The proposed method demonstrates a high level of accuracy in the identification of moldy-core apples, while also offering advantages in terms of simplicity, speed, cost-effectiveness and non-contact detection.

Research on Pig Sound Recognition Based on Deep Neural Network and Hidden Markov Models.

In order to solve the problem of low recognition accuracy of traditional pig sound recognition methods, deep neural network (DNN) and Hidden Markov Model (HMM) theory were used as the basis of pig sound signal recognition in this study. In this study, the sounds made by 10 landrace pigs during eating, estrus, howling, humming and panting were collected and preprocessed by Kalman filtering and an improved endpoint detection algorithm based on empirical mode decomposition-Teiger energy operator (EMD-TEO) cepstral distance. The extracted 39-dimensional mel-frequency cepstral coefficients (MFCCs) were then used as a dataset for network learning and recognition to build a DNN- and HMM-based sound recognition model for pig states. The results show that in the pig sound dataset, the recognition accuracy of DNN-HMM reaches 83%, which is 22% and 17% higher than that of the baseline models HMM and GMM-HMM, and possesses a better recognition effect. In a sub-dataset of the publicly available dataset AudioSet, DNN-HMM achieves a recognition accuracy of 79%, which is 8% and 4% higher than the classical models SVM and ResNet18, respectively, with better robustness.

Automated Detection of Rheumatic Heart Disease via Unsegmented Heart Sound Analysis

Rheumatic Heart Disease (RHD) remains a significant cause of morbidity and mortality in many parts of the world, particularly in low-resource settings. Early detection and intervention are critical to mitigating its progression and reducing associated complications. This paper presents an automated approach for the detection of RHD utilizing unsegmented heart sound analysis through deep learning techniques. By leveraging Spectro-temporal representations of raw heart sound signals, our proposed method aims to capture subtle yet discriminative patterns indicative of RHD pathology. The deep learning model is trained on a dataset comprising a diverse range of heart sounds, including those from both RHD-positive and RHD-negative individuals. Through rigorous evaluation of unseen data, our approach demonstrates promising performance in accurately distinguishing between RHD and healthy heart sounds. Furthermore, the proposed methodology offers the potential for scalable and cost-effective screening in resource-limited settings, thereby facilitating early identification and management of RHD cases. This work contributes to the advancement of computer-aided diagnostic tools for cardiovascular diseases, with implications for improving healthcare outcomes and reducing the burden of RHD globally. Key Words: Rheumatic Heart Disease, Heart Sound Analysis, Deep Learning, Automated Detection, Unsegmented Signals, Spectro-Temporal Representation, Cardiovascular Disease.

Interpretability of learning in a signal processing system

The paper presents a software package that allows one to generate algorithms for automatic classification of signals. The software package includes an algorithm that converts records of continuous signals into vector form, a set of machine learning methods, as well as data mining tools aimed at achieving transparency and interpretability of learning. The approach is based on the presentation of differences between compared classes as a set of relatively simple, statistically significant and interpretable effects, which are graphically represented on two-dimensional diagrams. The performance of the method is illustrated on the problem of assessing the state of the hive by sound signals. The software package can be used in solving applied problems of automatic diagnostics and data analysis.

DWSTr: a hybrid framework for ship-radiated noise recognition

The critical nature of passive ship-radiated noise recognition for military and economic security is well-established, yet its advancement faces significant obstacles due to the complex marine environment. The challenges include natural sound interference and signal distortion, complicating the extraction of key acoustic features and ship type identification. Addressing these issues, this study introduces DWSTr, a novel method combining a depthwise separable convolutional neural network with a Transformer architecture. This approach effectively isolates local acoustic features and captures global dependencies, enhancing robustness against environmental interferences and signal variability. Validated by experimental results on the ShipsEar dataset, DWSTr demonstrated a notable 96.5\% recognition accuracy, underscoring its efficacy in accurate ship classification amidst challenging conditions. The integration of these advanced neural architectures not only surmounts existing barriers in noise recognition but also offers computational efficiency for real-time analysis, marking a significant advancement in passive acoustic monitoring and its application in strategic and economic contexts.

Spectro Temporal Fusion with CLSTM-Autoencoder based approach for Anomalous Sound Detection

Deep learning models are proved efficient for complex learning tasks. Anomalous sound detection is one such complex task for which self-supervised deep architectures are emerging in recent days. Self-supervised deep models efficiently capture the underlying structure of data. Self-supervised anomalous sound detection attempts to distinguish between normal sounds and unidentified anomalous sounds. With the use of appropriate autoencoders, reconstruction error based decision making is effective for anomaly detection in domains such as computer vision. Auditory image (Spectrogram) based representation of sound signals are commonly used in sound event detection. We propose convolutional long short-term memory (CLSTM) Auto Encoder based approach for anomalous sound detection. In this approach, we explore fusion of spectral and temporal features to model characteristics of normal sounds with noises. The proposed approach is evaluated using MIMII dataset and the DCASE Challenge (2020) Task 2—Anomalous sound detection dataset. Experiments on proposed approach reveal significant improvement over the state-of-the-art approaches.

Elimination of heart sound from respiratory sound using adaptive variational mode decomposition for pulmonary diseases diagnosis

Lung sound (LS) signals are a vital source of information for the identification of pulmonary disorders. Heart sound (HS) is the most common contaminant of lung sounds during auscultation from the chest walls. This directly affects the efficiency of lung sound processing in diagnosing lung diseases. In this work, Adaptive Variational Mode Decomposition (AVMD) technique is proposed to remove heart sound contaminants from lung sounds. The proposed AVMD method initially breakdown the noisy lung sound signal into a collective of bandlimited modes called variational mode functions (VMF). Then, based on the frequency spectrum, the HS is filtered out from the LS. The real time lung sound data is collected from 95 participants and the performance of VMD technique is evaluated using the statistical metrics measures. Thus, the proposed topology exhibits Higher SNR (29.6587dB, lowest Root Mean Square (RMSE) of 0.0102, lowest normalized Mean Absolute Error (nMAE) of 0.0336, and highest percentage in correlation coefficient Factor (CCF) of 99.79% respectively. These experimental results are found to be superior and outperform all other recently proposed techniques.

CARgram: CNN-based accident recognition from road sounds through intensity-projected spectrogram analysis

Road surveillance systems play an important role in traffic monitoring and detecting hazardous events. In recent years, several artificial intelligence-based approaches have been proposed for this purpose, typically based on the analysis of the acquired video streams. However, occlusions, poor lighting conditions, and heterogeneity of the events may often reduce their effectiveness and reliability. To overcome the limitations mentioned, scientific and industrial research has therefore focused on integrating such solutions with audio recognition methods. By automatically identifying anomalous traffic sounds, e.g., car crashes and skids, they help reduce false positives and missed alarms. Following this trend, in this work, we propose an innovative pipeline for the analysis of intensity-projected audio spectrograms from streams of traffic sounds, which exploits both (i) a visual approach based on a custom, special-purpose Convolutional Neural Network for the identification of anomalous events on the sound signal; and, (ii) a novel multi-representational encoding of the input, which proved to significantly improve the recognition accuracy of the neural models. The validation results of the proposed pipeline on the public MIVIA dataset, with a 0.96% of false positive rate, showed to be the best performance against the state-of-the-art competitors. Notably, following such findings, a prototype implementation has been deployed on a real-world video surveillance infrastructure.

A Deep Learning-Based Embedded System for Pest Bird Sound Detection and Proximity Estimation

Cultivating crops is vital for driving economies, and maintaining agricultural fields is crucial for sustaining food production. This initiative centers on addressing the issue of pest birds, specifically starlings, within vineyards. The proposed strategy employs sound signals to detect and distinguish starling birds within the vineyard environment. Through an analysis of audio inputs from the surroundings, the system can effectively recognize unique sound patterns associated with starling birds, utilizing deep learning techniques. Furthermore, this project incorporates ultrasonic sensors for distance estimation, enabling the calculation of the bird’s proximity from a fixed point within the vineyard. All of these detection and estimation processes are executed on a RP2040 microcontroller, specifically the Cortex-M0+ 133 MHz variant. Following the detection phase, an autonomous vehicle equipped with red diode lasers can be dispatched to the designated location to deter the pest birds and safeguard the vineyards from unwanted disruptions and crop losses.

Noninvasive wearable sensor for the continuous monitoring of human sound and movement signals in real-time

Recently, with the development of non-invasive human health monitoring technology including wearable devices, a flexible sensor that monitors ‘human sound and movement signals’ such as human voice and muscle movement is attracting attention. In this experiment, electrospun nanofibers were mixed with highly conductive nanoparticles and coated with polyaniline to detect the patient's electrical signals. Due to the high piezoelectric effect, nanofiber-based sensors do not require charging through a separate battery, so they can be used as self-powered devices. In addition, the LCR meter test confirmed that the sensor has a high capacitance due to its high conductivity and high sensitivity to electrical signals. The sensor produced in this study can visually estimate the electrical signal of the actual human body through real-time comparison with electromyography (EMG) measuring equipment, and it was confirmed that the error is small. This sensor is expected to be widely used in the medical field, from simple sound and movement signals to disease monitoring.

Sound amplitude (dB) of male Western Capercaillie Tetrao urogallus

Knowledge of the true amplitude (dB) of sound signals is necessary for calculating how far they can travel and the distance at which they can be detected, two crucial aspects for planning bio-acoustic sampling protocols, particularly if endangered species such as the Western Capercaillie are involved. In order to be able to measure the true emission amplitude of male Western Capercaillie Tetrao urogallus calls, we recorded 90 repetitions of the call of a single male using a calibrated microphone located 1 m away from the bird. In this way, we were able to determine for the first time the true amplitude of a calling male Western Capercaillie in the Alt Pyrenees Natural Park, which turned out to be 67.68 ± 0.83 dB. Considering the characteristics of our microphone, this gives an average maximum detection distance of 174.8 m, important information for optimising future bio-acoustic detection and monitoring studies.