Search Results (58)

Despite having approximately 24 million native speakers, Romanian remains a low-resource language for automatic speech recognition (ASR), with few accurate and publicly available systems. To address this gap, this study explores the challenges of adapting modern speech recognition models, such as wav2vec 2.0 and Conformer, to Romanian. Our investigation is a comprehensive analysis of the two models, their capabilities to adapt to Romanian data, and the performance of the trained models. The research also focuses on unique attributes of the Romanian language, data collection techniques, including weakly supervised learning, and processing methodologies. Building on the previously introduced Echo dataset of 378 h, we release CRoWL (Crawled Romanian Weakly Labeled), a weakly supervised dataset of 9000 h created via automatic transcription. We obtain strong results that, to the best of our knowledge, are competitive with or exceed publicly reported results for Romanian under comparable open evaluation settings, with Conformer attaining 3.01% WER on Echo + CRoWL and wav2vec 2.0 reaching 4.04% (Echo) and 4.17% (Echo + CRoWL). In addition to the datasets, we also release our most capable models as open source, along with their training plans, thereby providing a solid foundation for researchers interested in languages with limited representation. Full article

Z-cut lithium niobate single crystal demonstrates considerable promise for contact-based ultrasonic nondestructive testing and structural health monitoring (SHM) transducers due to its high piezoelectric coefficients, strong electromechanical coupling capability, and environmentally friendly lead-free composition. As a simulation-based theoretical exploration, this study systematically investigates the impact of gap spacing and electrode width in concentric ring configurations on the resonant characteristics and pulse-echo response of ultrasonic transducers by establishing a parametrized finite element model. Numerical simulations reveal that electrode geometry plays a critical role in determining both the effective electromechanical coupling coefficient and echo signal strength. Optimizing the electrode ring width achieved an effective electromechanical coupling coefficient ($k_{\mathit{eff}}$) of 35.2%, while systematic enlargement of the electrode gap further enhanced this value to 50.8%. The study also demonstrates that optimized ring width and adjusted electrode spacing increased the echo signal’s peak-to-peak amplitude ($V_{pp}$) by factors of 4.94 and 2.03, respectively, compared to the poorest-performing configuration within each parameter group. This study establishes that precise design of concentric electrode configurations serves as an effective strategy for tuning lithium niobate ultrasonic transducer characteristics, providing critical design guidelines for developing high-performance ultrasonic transducers for solid medium coupling. Full article

Background and Aims: Solid pseudopapillary neoplasms (SPN) are rare neoplasms of the pancreatic gland. Despite the indolent behavior, surgical resection is required according to the risk of metastasis development. Few data are available on endoscopic ultrasound (EUS) features, comparison between the features described at EUS and contrast-enhanced, and diagnostic accuracy and safety of EUS-guided FNB in these tumors. Patients and Methods: All consecutive patients with a EUS-guided FNB-based pathological diagnosis of SPN were extrapolated from a prospectively maintained database. Demographic, radiologic, and echo-endoscopic features were collected. FNB specimens were re-evaluated from two expert pathologist and the main histological features of SPN were investigated. Results: Thirty-seven patients were included (32 females and 5 males), with a mean age of 35.8 ± 15.8 years. Contrast-enhanced imaging based diagnosis was accurate in 20 patients (54.1%). EUS features were significantly different compared to contrast-enhanced imaging in terms of cystic appearance (40.5% vs. 16.2%; p = 0.03) and vascular pattern (p = 0.01). FNB-based diagnosis of SPN was confirmed on surgical specimen in all, 37 patients, resulting in a diagnostic accuracy of EUS-guided FNB of 100%. Only one patient (2.6%) experienced a mild procedure-related adverse event. Discussion: Contrast-enhanced imaging based diagnosis of SPN is difficult. Despite the rarity of the disease, EUS-guided FNB with 22-gauge fork-tip needle has a very high diagnostic accuracy for SPN, with rare and mild adverse events. Full article

Triarylmethyl radicals (TAMs) have recently emerged as highly effective polarizing agents in dynamic nuclear polarization (DNP) under viscous conditions, enabling substantial hyperpolarization via the solid-effect (SE) DNP mechanism even at room temperature. A comparable, though less pronounced, enhancement was observed for BDPA radicals embedded in phosphocholine-based lipid bilayers. Given the increasing interest in elucidating the structure and dynamics of biopolymers and their high-molecular-weight assemblies—such as cell membranes—this study focuses on the design, synthesis, and characterization of paramagnetic agents tailored for DNP-based structural biology. To this end, we synthesized a series of TAM derivatives functionalized with lipophilic substituents and characterized their magnetic resonance properties, including isotropic hyperfine interaction (HFI) constants on carbon nuclei and electron spin relaxation times (T₁ and T_m) at low temperatures (80 K). Echo-detected EPR spectra and electron spin echo envelope modulations (ESEEM) were recorded for novel TAM incorporated into liposomes composed of 1,2-dioleoyl-sn-glycero-3-phosphocholine (DOPC). These low-temperature measurements revealed that the radicals are localized either at the liposome surface or within the lipid bilayer, ensuring optimal accessibility to water molecules. Crucially, the presence of a single cholesterol moiety provides strong noncovalent anchoring within the hydrophobic core of the bilayer. Guided by these findings, we identify an amphiphilic TAM bearing a single cholesterol group and polar carboxyl functionalities as a highly promising candidate for DNP applications in membrane biology, combining efficient polarization transfer, bilayer integration, and aqueous accessibility. Full article

There are various non-destructive techniques for determining the internal properties of materials in fluids, semi-solids, solids, and biological tissue. One of these techniques is low-intensity ultrasonic testing. In this proceeding paper, a study on the architecture of a piezoelectric acoustic detector (PAD) is presented, from which an analysis for the design, development, and construction of an acoustic wave detector in the ultrasonic spectrum has emerged. Its purpose is to be applied to soft matter and tissue. The 110 μm thick polyvinylidene fluoride (PVDF) piezoelectric element was used as the active element in the thickness mode configuration. Piezoelectric constitutive equations were applied to a one-dimensional model for the analysis. A cylindrical iron–nickel backing was used, and the parts were bonded with conductive silver epoxy glue. The results are presented. The equation for the output voltage of the piezoelectric acoustic detector is described. Functional testing of the PAD is demonstrated using the pulse-echo technique, in which an acoustic wave generator excites an ultrasonic immersion sensor in emission configuration and the DAP is connected to a digital oscilloscope to observe the received signal. Finally, pulsed photoacoustic spectroscopy was applied to a biological tissue emulator and yielded significant results in the detection of a ruby sphere embedded in the emulator. It is proposed to further investigation the DAP models in multilayer structural configurations to increase their sensitivity. Full article

This paper presents a study on the nonlinear vibrations in the impact-echo (IE) method for void flaw detection of solid structures. Linear theory has historically served as the foundational framework for non-destructive methods, including the IE method, particularly for estimating flaws in solids. This paper gives a comprehensive analysis of the nonlinear theory behind the IE method for detection of voids in solids such as concrete structures. The general equation of motion is presented for the flexural vibration of a void-defected solid with general nonlinear constitutive material properties, and then the simplified solutions for polynomial nonlinearity and hysteresis nonlinearity are derived comprehensively. The solutions of principal frequency and sub- and super-harmonics as well as the frequency of combined modes are elaborated, and the theoretical formula of resonant frequency shift with amplitude is derived. As conventional nonlinear IE methods have been conducted by only using a phenomenological model of linear shift in resonant frequency with amplitude, the proposed new frame of nonlinear vibration theory can be used to implement the IE method more comprehensively and accurately for void detection in solids. Full article

Ultrasonic measurement techniques are increasingly used to measure the burning rates of solid rocket fuel, but challenges arise due to noise and signal attenuation caused by the motor’s multi-layered structure. This paper proposes an adaptive thresholding method combined with a wavelet threshold function for effective ultrasonic signal denoising. Additionally, an extreme value feature fitting algorithm is introduced for accurate echo signal localization, even in low signal-to-noise ratio (SNR) conditions. Numerical simulations show a 10 dB improvement in SNR at −20 dB, with a correlation coefficient of 0.83 between the denoised and true signals. Echo localization tests across 12 SNR levels demonstrate a consistent error below 1 μs. Compared to other algorithms, the proposed method achieves higher precision, with a maximum displacement error of 0.74 mm. Hardware-in-the-loop experiments show an increase in SNR from −15 dB to 5.78 dB, with maximum displacement and rate errors of 0.9239 mm and 0.781 mm/s. In fuel-burning experiments, the burning rate curve closely matches the theoretical curve, with an initial fuel thickness error of only 0.12 mm, confirming the method’s effectiveness in complex environments. Full article

This research presents a “flexible support structure between reflective mirrors” through a coupling analysis method to restrain the surface shaping error of reflectors in the optical system of airborne LiDAR bathymetry (ALB) under various working conditions. The flexible structure proposed adjusts the mechanical relationship between the reflectors and the support structure to reduce reflector mirror deformation. The optical system is first modeled using Zemax and exported to SolidWorks to create a 3D model of the optical receiving system. Ansys is then used to conduct stiffness testing and surface analysis on the support structure of the annular thin cylinder. According to the analysis results, the first-order frequency of the support structure using a ring-shaped thin cylinder is as high as 353.64 Hz, which indicates that it has good dynamic characteristics. The PV value of the reflector mirror deformation under the thermal coupling reaches 32.59 nm, and the RMS value reaches 8.62 nm. Additionally, it is discovered that the maximum acceleration response of the reflector mirror under the applied 1 g acceleration excitation reaches 4.22 g when carrying out the dynamics analysis of the support structure. Under random vibration analysis, the maximum acceleration RMS value of the reflector mirror assembly reaches 2.18 g, and the maximum stress of the flexible device of the support structure reaches 2.65 MPa. Especially, five groups of experimental results demonstrated that the proposed coupling analysis method can receive the echo signals, the reflector mirror support structure designed in this paper, and the flexible structure is stable and reliable. Full article

Advanced ¹H Nuclear Magnetic Resonance (NMR) relaxometry and diffusometry methods and VIS-nearIR spectroscopy combined with pH, electrical conductivity (EC) and totally dissolved solids (TDSSs) measurements were used to assess the properties of wastewater collected from a chicken slaughterhouse in each step of the treatment process (wastewater before treatment, biologically treated wastewater, chemically treated wastewater and discharged wastewater) and from sludge. The ¹H NMR Carr–Purcell–Meiboom–Gill (CPMG) and Pulsed-Gradient-Stimulated-Echo (PGSE) decay curves recorded for all samples of wastewater were analyzed by inverse Laplace transform (ILT) to obtain the distributions of transverse relaxation times T₂ and diffusion coefficient D. The VIS-nearIR total absorbance, T₂-values, D-values, pH, EC and TDSS parameters were used for statistical analysis in principal component (PCA). The ¹H T₂-distributions measured for the slaughterhouse wastewater lie in two main regions reflecting the number of dissolved solids or the distribution of undissolved solids. The PCA analysis successfully differentiates between polluted and less polluted wastewaters and sludge. The wastewater treatment applied by the slaughterhouse is efficient. The recommended methods for wastewater monitoring are the NMR T₂- and D-distributions and EC, TDSSs and NMR-D diffusion coefficient. Finally, Machine Learning algorithms are used to provide prediction maps of wastewater treatment stage. Full article

Friedrich Hölderlin’s late hymns Patmos (first version) and Mnemosyne (early draft) create an intriguing tension between the “solid letter” that must be deciphered faithfully and the evocation of a “sounding script” that, together with an equally enigmatic “echo”, refuses direct hermeneutic understanding. At the point where the reader’s interpretive desire threatens to fail, musical settings like Peter Ruzicka’s MNEMOSYNE: Remembrance and Forgetting can be listened to as an attempt to actualize what Hölderlin’s original writing must leave unrealizable: the presence of real sound. In this audio-hermeneutic transfer, the act of listening opens up possibilities of the audible that are promised by the literary text without being actualized. The present essay interrogates this intermedial translatability between letter and sound by isolating a few selected passages from the facsimile reproduction of Hölderlin’s palimpsestic manuscript of multiple revisions, as provided by the Frankfurter Ausgabe. Mindful of the discontinuities and gaps in the original poems, my own analysis foregrounds its own fragmentary mode of reading Hölderlin’s poetry and listening to Ruzicka’s music. Full article

Carrier-free ultra-wideband sensors have high penetrability anti-jamming solid ability, which is not easily affected by the external environment, such as weather. Also, it has good performance in the complex jungle environment. In this paper, we propose a jungle vehicle identification system based on a carrier-free ultra-wideband sensor. Firstly, a composite jungle environment with the target vehicle is modeled. From this model, the simulation obtains time-domain echoes under the excitation of carrier-free ultra-wideband sensor signals in different orientations. Secondly, the time-domain signals are transformed into MTF images through the Markov transfer field to show the statistical characteristics of the time-domain echoes. At the same time, we propose an improved RepVGG network. The structure of the RepVGG network contains five stages, which consist of several RepVGG Blocks. Each RepVGG Block is created by combining convolutional kernels of different sizes using a weighted sum. We add the self-attention module to the output of stage 0 to improve the ability to extract the features of the MTF map and better capture the complex relationship between characteristics during training. In addition, a self-attention module is added before the linear layer classification output in stage 4 to improve the classification accuracy of the network. Moreover, a combined cross-entropy loss and sparsity penalty loss function helps enhance the performance and accuracy of the network. The experimental results show that the system can recognize jungle vehicle targets well. Full article

Ultrasonic nondestructive testing (NDT) usually utilizes conventional bulk piezoelectric transducers as transceivers. However, the complicated preparation and assembly process of bulk piezoelectric ceramics limits the development of NDT probes toward miniaturization and high frequency. In this paper, a 4.4 mm × 4.4 mm aluminum nitride (AlN) piezoelectric micromachined ultrasonic transducer (PMUT) array is designed, fabricated, characterized, and packaged for ultrasonic pulse–echo NDT of solids for the first time. The PMUT array is prepared based on the cavity silicon-on-insulator (CSOI) process and packaged using polyurethane (PU) material with acoustic properties similar to water. The fabricated PMUT array resonates at 2.183 MHz in air and at around 1.25 MHz after PU encapsulation. The bandwidth of the packaged PMUT receiver (244 kHz) is wider than that of a bulk piezoelectric transducer (179 kHz), which is good for axis resolution improvement. In this work, a hybrid ultrasonic NDT probe is designed using two packaged PMUT receivers and one 1.25 MHz bulk transmitter. The bulk transmitter radiates an ultrasonic wave into the sample, and the defect echo is received by two PMUT receivers. The 2D position of the defect could be figured out by time-of-flight (TOF) difference, and a 30 mm × 65 mm detection area is acquired. This work demonstrates the feasibility of applying AlN PMUTs to ultrasonic NDT of solids and paves the way toward a miniaturized NDT probe using AlN PMUT technology. Full article

Investigating the dynamic evolution process of the ocean and ionosphere in sudden sea conditions poses a challenging problem. To address this objective, this study utilizes actual data from high-frequency surface wave radar (HFSWR) to analyze, validate, summarize, and characterize the echo properties of the ocean and ionosphere during the severe Typhoon Muifa. By employing the short-time Fourier transform (STFT) method, the HFSWR ocean and ionosphere echoes stimulated by typhoon-induced gravity waves are observed, and the joint gravity wave features of the ocean and ionosphere echoes at different time scales are extracted. Additionally, the phase-space reconstruction method is employed to characterize the dynamical evolution of the joint gravity wave features in higher-dimensional space. Furthermore, the chaotic dynamics behavior of the joint gravity wave features is analyzed using the largest Lyapunov exponents. By combining the gravity wave features with chaotic dynamics, this study introduces a method to characterize the joint gravity wave features. The extraction of joint gravity wave features in HFSWR echoes stimulated by typhoons, along with the construction of a chaotic characterization scheme for the gravity wave features, provides an innovative approach and a solid technical foundation for studying the ocean and ionosphere using HFSWR under sudden sea conditions. Full article

High-methoxyl apple pectin (AP) derived from apple was employed as the main ingredient facilitating rheological modification features in developing dysphagia-friendly fluidized alimentary matrices. Xanthan gum (XG) was also included as a composite counterpart to modify the viscoelastic properties of the thickened system under different thermal processes. The results indicate that AP is extremely sensitive to thermal processing, and the viscosity is greatly depleted under a neutral pH level. Moreover, the inclusion of calcium ions echoed the modification effect on the rheological properties of AP, and both the elastic property and viscosity value were promoted after thermal processing. The modification effect of viscoelastic properties (G′ and G″) was observed whne XG was incorporated into the composite formula. Increasing the XG ratio from 7:3 to 6:4 (AP:XG) triggers the rheological transformation from a liquid-like form to a solid-like state, and the viscosity value shows that the AP-XG composite system exhibits better thermal stability after thermal processing. The ambient modifiers of pH (pH < 4) and calcium chloride concentration (7.5%) with an optimal AP-XG ratio of 7:3 led to weak-gel-like behavior (G″ < G′), helping to maintain the texture properties of dysphagia-friendly features similar to those prior to the thermal processing. Full article

Solid-state NMR (ss-NMR) is a powerful tool to investigate noncrystallizable, poorly soluble molecular systems, such as membrane proteins, amyloids, and cell walls, in environments that closely resemble their physical sites of action. Rotational-echo double resonance (REDOR) is an ss-NMR methodology, which by reintroducing heteronuclear dipolar coupling under magic angle spinning conditions provides intramolecular and intermolecular distance restraints at the atomic level. In addition, REDOR can be exploited as a selection tool to filter spectra based on dipolar couplings. Used extensively as a spectroscopic ruler between isolated spins in site-specifically labeled systems and more recently as a building block in multidimensional ss-NMR pulse sequences allowing the simultaneous measurement of multiple distances, REDOR yields atomic-scale information on the structure and interaction of proteins. By extending REDOR to the determination of ¹H–X dipolar couplings in recent years, the limit of measurable distances has reached ~15–20 Å, making it an attractive method of choice for the study of complex biomolecular assemblies. Following a methodological introduction including the most recent implementations, examples are discussed to illustrate the versatility of REDOR in the study of biological systems. Full article