Hasil untuk "Acoustics. Sound"

G. Griffiths, W. Schiesser

The study of waves can be traced back to antiquity where philosophers, such as Pythagoras (c.560-480 BC), studied the relation of pitch and length of string in musical instruments. However, it was not until the work of Giovani Benedetti (1530-90), Isaac Beeckman (1588-1637) and Galileo (1564-1642) that the relationship between pitch and frequency was discovered. This started the science of acoustics, a term coined by Joseph Sauveur (1653-1716) who showed that strings can vibrate simultaneously at a fundamental frequency and at integral multiples that he called harmonics. Isaac Newton (1642-1727) was the first to calculate the speed of sound in his Principia. However, he assumed isothermal conditions so his value was too low compared with measured values. This discrepancy was resolved by Laplace (1749-1827) when he included adiabatic heating and cooling effects. The first analytical solution for a vibrating string was given by Brook Taylor (1685-1731). After this, advances were made by Daniel Bernoulli (1700-82), Leonard Euler (1707-83) and Jean d’Alembert (1717-83) who found the first solution to the linear wave equation, see section (3.2). Whilst others had shown that a wave can be represented as a sum of simple harmonic oscillations, it was Joseph Fourier (1768-1830) who conjectured that arbitrary functions can be represented by the superposition of an infinite sum of sines and cosines now known as the Fourier series. However, whilst his conjecture was controversial and not widely accepted at the time, Dirichlet subsequently provided a proof, in 1828, that all functions satisfying Dirichlet’s conditions (i.e. non-pathological piecewise continuous) could be represented by a convergent Fourier series. Finally, the subject of classical acoustics was laid down and presented as a coherent whole by John William Strutt (Lord Rayleigh, 1832-1901) in his treatise Theory of Sound. The science of modern acoustics has now moved into such diverse areas as sonar, auditoria, electronic amplifiers, etc.

D. Begault

M. S. Howe

E. Simmonds, D. MacLennan

F. Jensen, W. Kuperman, M. Porter et al.

A. M. Bratkovsky

Nematic aerogel (nAG) supports so-called polar phase in liquid 3He. The experiment [Dmitriev et al, JETP Lett. 112, 780 (2020)] showed that the onset of polar phase inside the nAG is accompanied by emergence of a sound wave with frequency quickly growing with cooling down from transition temperature and reaching a plateau. To describe this behavior, we start by calculating the elastic properties of the dry nematic AG that appear to depend only on Young's modulus of the parent material (e.g. mullite), the volume fraction of the solid phase and the aspect ratio of the representative volume of nAG. The elastic constants are then used to solve elasto-hydrodynamic equations for various sound vibrations of nAG filled with 3He. The (isotropic) first sound and anisotropic second sound in the polar phase are strongly hybridized with fourth sound and standard elastic modes in nAG. The hybrid second and the transverse fourth sound start with zero velocity at the transition, similar to pure 3He, and quickly grow with lowering temperature until they hit the sample finite size cutoff.

Jorge L. Mejía-Méndez, Edwin E. Reza-Zaldívar, Diego E. Navarro-López et al.

This study introduces a novel combination of sonochemical synthesis and machine learning (ML) modeling to analyze the effect of Lanthanum (La) doping on the biological properties of cerium oxide (CeO2) nanoparticles with various La concentrations (0, 1, 5, and 10 at.%). Ultrasonic-assisted synthesis enabled La incorporation into the CeO2 lattice, resulting changes in crystallinity, lattice parameters, and surface features. Detailed characterization confirmed successful doping and indicated stable nanoparticle suspensions with controlled size ranges. Sonochemical synthesis promoted the oxidation of Ce3+ to Ce4+. Biological testing revealed low cytotoxicity across various cell lines (HepG-2, 3 T3-L1, Caco-2, and U87) and increased antioxidant activity, especially in samples with 5 and 10 at.% La, which demonstrated improved free radical scavenging of DPPH and H2O2 radicals. Notably, advanced ML models—including Extremely Randomized Trees, random forest, Gradient Boosting, and LightGBM—enabled the prediction of antioxidant activity based on nanoparticle features, identifying antioxidant method, concentration, and chemical composition as key factors influencing biological effects. This combined experimental and data-driven approach not only clarifies the structure–activity relationships of La-doped CeO2 nanoparticles but also emphasizes the significant potential of ML in designing and optimizing nanomaterials for biomedical applications. The combination of sonochemical synthesis and ML modeling provides a robust framework for accelerating nanomaterial development by minimizing trial-and-error experiments and providing mechanistic insights into their biological functions.

Dayun Choi, Jung-Woo Choi

Recent advances in target sound extraction (TSE) utilize directional clues derived from direction of arrival (DoA), which represent an inherent spatial property of sound available in any acoustic scene. However, previous DoA-based methods rely on hand-crafted features or discrete encodings, which lose fine-grained spatial information and limit adaptability. We propose SoundCompass, an effective directional clue integration framework centered on a Spectral Pairwise INteraction (SPIN) module that captures cross-channel spatial correlations in the complex spectrogram domain to preserve full spatial information in multichannel signals. The input feature expressed in terms of spatial correlations is fused with a DoA clue represented as spherical harmonics (SH) encoding. The fusion is carried out across overlapping frequency subbands, inheriting the benefits reported in the previous band-split architectures. We also incorporate the iterative refinement strategy, chain-of-inference (CoI), in the TSE framework, which recursively fuses DoA with sound event activation estimated from the previous inference stage. Experiments demonstrate that SoundCompass, combining SPIN, SH embedding, and CoI, robustly extracts target sources across diverse signal classes and spatial configurations.

Chaoyue Niu, Veronica Rowe, Guy J. Brown et al.

Paediatric obstructive sleep apnoea (OSA) is clinically significant yet difficult to diagnose, as children poorly tolerate sensor-based polysomnography. Acoustic monitoring provides a non-invasive alternative for home-based OSA screening, but limited paediatric data hinders the development of robust deep learning approaches. This paper proposes a transfer learning framework that adapts acoustic models pretrained on adult sleep data to paediatric OSA detection, incorporating SpO2-based desaturation patterns to enhance model training. Using a large adult sleep dataset (157 nights) and a smaller paediatric dataset (15 nights), we systematically evaluate (i) single- versus multi-task learning, (ii) encoder freezing versus full fine-tuning, and (iii) the impact of delaying SpO2 labels to better align them with the acoustics and capture physiologically meaningful features. Results show that fine-tuning with SpO2 integration consistently improves paediatric OSA detection compared with baseline models without adaptation. These findings demonstrate the feasibility of transfer learning for home-based OSA screening in children and offer its potential clinical value for early diagnosis.

Terese von Knorring, Tobias Buhl Ihlemann, Paul Blanche et al.

Photoacoustic imaging (PAI) shows promise for skin cancer diagnosis by detecting chromophores like melanin, hemoglobin, lipids, and collagen. While most studies focus on malignant lesions, understanding normal skin variability across anatomical regions is crucial for validating PAI's clinical application and its use in melanoma diagnosis. We assessed normal skin in 20 healthy volunteers from three different body locations using a clinical PAI system and compared suspicious looking pigmented skin lesions, including melanomas, to adjacent normal skin (n = 74). Higher deoxyhemoglobin levels were observed in the ankle compared to the cheek and volar forearm, while melanin, lipids, and collagen showed minimal variation. Patients with malignant lesions had significantly higher deoxyhemoglobin levels (p = 0.001) than adjacent normal skin, a difference not seen in benign lesions. These findings suggest that PAI may help diagnose malignancies by identifying increased vascularity in skin cancers, while anatomical differences should be considered during diagnostic work-up.

Chaoping Gui

Refrigerators bring convenience to people, but the noise they produce can also affect people’s lives. Refrigerator noise is dominated by low-frequency noise, and the active noise control method is more effective for this kind of noise. However, the existing active noise control methods for refrigerators do not take into account factors such as the limited space and complex structure of the refrigerator compressor room, and external interference signals will be introduced to affect the noise reduction performance during the error signal collection. Given that, this paper proposes the cascade notch feedback algorithm, which can well reduce the influence of external interference signals to better achieve the refrigerator noise reduction. The algorithm consists of the cascaded notch filtering algorithm and the feedback algorithm. The cascade notch filtering algorithm is formed by cascading the adaptive notch filtering algorithm, which is used to deal with the main frequency and harmonic noise generated by the refrigerator. The feedback algorithm consists of a robust algorithm for dealing with external interference signals introduced during error signal collection. The simulation experiment proves that the algorithm has advantages in terms of noise reduction and computational cost compared with the adaptive notch filtering algorithm and the improved algorithm. The experimental test platform is set up to carry out the actual refrigerator noise reduction experiments under different conditions. The algorithm has the effect of noise reduction under different robust algorithms.

Aditya Kommineni, Rajat Hebbar, Sarah Petrosyan et al.

Speech foundation models (SFMs) achieve state-of-the-art results in many tasks, but their performance on elderly, multilingual speech remains underexplored. In this work, we investigate SFMs' ability to analyze multilingual speech from older adults using spoken language identification as a proxy task. We propose three key qualities for foundation models to serve multilingual aging populations: robustness to input duration, invariance to speaker demographics, and few-shot transferability in low-resource settings. Zero-shot evaluation indicates a noticeable performance drop for shorter inputs. We find that native speakers' speech consistently outperforms non-native speech across languages. Few-shot learning indicates better transferability in larger models.

Haixin Zhao, Nilesh Madhu

Abstract Numerous post-processing methods have been proposed to improve coded speech quality and intelligibility. However, achieving state-of-the-art enhancement and generalisation across varying distortion levels remains a challenge. To bridge this gap, we propose a Lightweight Causal-Transformer-based Coded Speech Enhancement (LCT-CSE) model employing a causal frequency-time-frequency (FTF) transformer block. This block facilitates temporal and spectral sequential modelling using transformers, efficiently exploiting global dependency across causal-context TF bins while minimising computational overhead. Experimental results indicate that the proposed LCT-CSE model outperforms the considered baselines across mainstream lossy audio codecs, including Opus, AMR-WB, EVS and LC3+, with less footprint and complexity. To further utilise auxiliary, utterance-level information such as bitrate and other general distortion characteristics, building upon the LCT-CSE model, we propose two information incorporation methods. One employs one-hot vector representations and feature fusions, referred to as 1-hot vector-based modulation, while the other dynamically switches information-dependent network paths, termed dynamic linear modulation (DLM). These methods can be used to improve performance in bitrate-information utilisation, with negligible additional computational overhead. The DLM model even achieves comparable performance to bitrate-specific trained (BST) models. We further extend the proposed information incorporation method, DLM, to a generalised scenario, tandem coding. Compared to the two practically used approaches, the DLM-based LCT-CSE model consistently exhibits improved generalisability across varying tandem encoding conditions, based on derivative distortion information. Specifically, it achieves gains up to 0.74 in PESQ, 7% in STOI, and 0.18 in MOS-SIG under various bitrate conditions. This indicates significant potential for further applications where auxiliary information can be utilised.

Joseph Carlo U. Candare, Kristian Hauser A. Villegas

The sound modes of a flowing superfluid is described by the massless Klein-Gordon equation in an effective background metric. This effective background metric can be designed to mimick a black hole using the acoustic horizon. In this work, we study the AdS/CFT dual of the sound modes in the presence of an acoustic horizon in the bulk. Focusing on fluids with a purely radial flow, we derive the metric tensor for the effective acoustic spacetime and deduce a necessary condition for an acoustic black hole geometry to exist within the fluid. Using specific examples of superfluid velocity profiles, we obtained the source, operator expectation value, Green's function, and spectral density of the dual field theory by solving for the asymptotic behavior of the sound modes near the AdS boundary. In all our examples, the sound modes remain gapless but the excitations are described by branch cuts, instead of poles, which is typical of strongly coupled systems. Furthermore, we calculate the effective Hawking temperature of the dual field theory associated with the bulk acoustic horizon. Lastly, we investigate the near horizon properties and derive the superfluid velocity profile that can give rise to an infrared emergent quantum criticality.

June-Woo Kim, Miika Toikkanen, Yera Choi et al.

Respiratory sound classification (RSC) is challenging due to varied acoustic signatures, primarily influenced by patient demographics and recording environments. To address this issue, we introduce a text-audio multimodal model that utilizes metadata of respiratory sounds, which provides useful complementary information for RSC. Specifically, we fine-tune a pretrained text-audio multimodal model using free-text descriptions derived from the sound samples' metadata which includes the gender and age of patients, type of recording devices, and recording location on the patient's body. Our method achieves state-of-the-art performance on the ICBHI dataset, surpassing the previous best result by a notable margin of 1.17%. This result validates the effectiveness of leveraging metadata and respiratory sound samples in enhancing RSC performance. Additionally, we investigate the model performance in the case where metadata is partially unavailable, which may occur in real-world clinical setting.

Shoichi Koyama, Juliano G. C. Ribeiro, Tomohiko Nakamura et al.

The area of study concerning the estimation of spatial sound, i.e., the distribution of a physical quantity of sound such as acoustic pressure, is called sound field estimation, which is the basis for various applied technologies related to spatial audio processing. The sound field estimation problem is formulated as a function interpolation problem in machine learning in a simplified scenario. However, high estimation performance cannot be expected by simply applying general interpolation techniques that rely only on data. The physical properties of sound fields are useful a priori information, and it is considered extremely important to incorporate them into the estimation. In this article, we introduce the fundamentals of physics-informed machine learning (PIML) for sound field estimation and overview current PIML-based sound field estimation methods.

Rui Yang, Jiangnan Hu, Jiaqi Ding et al.

Betanin, a water-soluble colorant, is sensitive to light and temperature and is easily faded and inactivated. This study investigated the formation of yeast protein-chitooligosaccharide-betanin complex (YCB) induced by ultrasound treatment, and evaluated its protective effect on the colorant betanin. Ultrasound (200–600 W) increased the surface hydrophobicity and solubility of yeast protein, and influenced the protein’s secondary structure by decreasing the α-helix content and increasing the contents of β-sheet and random coil. The ultrasound treatment (200 W, 15 min) facilitated binding of chitooligosaccharide and betanin to the protein, with the binding numbers of 4.26 ± 0.51 and 0.61 ± 0.06, and the binding constant of (2.73 ± 0.25) × 105 M−1 and (3.92 ± 0.10) × 104 M−1, respectively. YCB could remain the typical color of betanin, and led to a smaller and disordered granule morphology. Moreover, YCB exhibited enhanced thermal-, light-, and metal irons (ferric and copper ions) -stabilities of betanin, protected the betanin against color fading, and realized a controlled release in simulated gastrointestinal tract. This study extends the potential application of the fungal proteins for stabilizing bioactive molecules.

Yue Zou, Yike Yang, Dong Han

The present study investigated the effects of language dominance on the cross-linguistic influence in the first and second languages (L1 and L2) of lexical tone production by Mandarin–Cantonese late bilinguals. Although the participants were unable to retain their L1 tonal system or to fully acquire the L2 tonal system after long-term exposure to their L2, certain correlations emerged between language dominance and tone production in L1 and L2. These findings add to the existing literature on language dominance and support the general assumption that bilinguals' two languages interfere with each other.

Yisha Xie, Kangyu Zhao, Feng Yang et al.

The hardness properties of unwashed surimi gel are considered as the qualities of gelation defect. This research investigated the effect of ultrasound-assisted first-stage thermal treatment (UATT) on the physicochemical properties of unwashed Silver Carp surimi gel, and the enhancement mechanism. UATT could reduce protein particle size, which significantly reduced from 142.22 μm to 106.70 μm after 30 min of UATT compared with the nature protein. This phenomenon can promote the protein crosslinking, resulting in the hardness of surimi gel increased by 15.08 %. Partially unfolded structure of myofibrillar protein and exposures of tryptophan to water, lead to the increase in the zeta potential absolute value, driven by UATT. The reduced SH group level and the conformational conversion of proteins from random coiling to α-helix and β-sheet, which was in support of intermolecular interaction and gel network construction. The results are valuable for processing protein gels and other food products.

Tanmay Khandelwal, Rohan Kumar Das

Sound event detection (SED) entails identifying the type of sound and estimating its temporal boundaries from acoustic signals. These events are uniquely characterized by their spatio-temporal features, which are determined by the way they are produced. In this study, we leverage some distinctive high-level acoustic characteristics of various sound events to assist the SED model training, without requiring additional labeled data. Specifically, we use the DCASE Task 4 2022 dataset and categorize the 10 classes into four subcategories based on their high-level acoustic characteristics. We then introduce a novel multi-task learning framework that jointly trains the SED and high-level acoustic characteristics classification tasks, using shared layers and weighted loss. Our method significantly improves the performance of the SED system, achieving a 36.3% improvement in terms of the polyphonic sound event detection score compared to the baseline on the DCASE 2022 Task 4 validation set.