Underwater acoustic signal denoising technology aims to overcome the challenge of recovering valuable ship target signals from noisy audios by suppressing underwater background noise. Traditional statistical-based techniques are difficult be applied effectively in complex environments, especially case extremely low signal-to-noise ratios (SNRs). To address these problems, we propose a noise-aware deep learning model with fullband-subband attention network (NAFSA-Net) for denoising. NAFSA-Net...

The remote sensing-based inversion of sound speed profile (SSP) enables the acquisition high-spatial-resolution SSP without in situ measurements. spatial division grid is crucial for accuracy results, determining both number samples and consistency relationships. result our research introduction a physics-inspired self-organizing map (PISOM) that facilitates by clustering according to physical perturbation law. linear relationship between sea surface parameters drives dimensionality...

The sonar image has limitations on the physical spatial resolution due to system configuration and underwater environment, which often leads challenges for targets detection. Here, deep learning method is applied enhance of images. Specifically, U-shaped end-to-end neural network contains down-sampling up-sampling parts proposed improve limited by array aperture. single target multiple cases are considered separately. In both cases, normalized loss testing sets declines rapidly, predicted...

Coastal acoustic tomography (CAT) is a remote sensing technique that utilizes methodologies to measure the dynamic characteristics of ocean in expansive marine domains. This approach leverages speed sound propagation derive vital parameters such as temperature and salinity by inversely estimating ray during its traversal through aquatic medium. Concurrently, analyzing different waves their round-trip enables inverse estimation hydrographic features, including flow velocity directional...

Underwater acoustic signal denoising is a challenging task due to the complexity of underwater environment. Most existing methods cannot effectively cope with problem (UWAS) at low signal-to-noise ratios (SNRs). According characteristics UWAS, novel idea proposed simultaneously model latent features from both time and frequency dimensions complex-valued spectrum in dual-branch self-attention network, namely DBSA-Net. In this model, magnitude phase information complex are enhanced different...

Research on ocean ambient noise is highly important for environment monitoring, marine mammal protection, underwater communication and navigation. In this paper, we present the long-term statistics wind dependence of near-bottom deep-sea in northwest South China Sea, at a depth 1240 m. The data were collected from 11 th July 2022 to 31 st December together with local speeds ranging 1 58 knots (two typhoons involved), processing frequency band between 20 2000 Hz. mean level calculated along...

BACE1 has been regarded as an essential drug design target for treating Alzheimer's disease (AD). Multiple independent Gaussian accelerated molecular dynamics simulations (GaMD), deep learning (DL), and mechanics general Born surface area (MM-GBSA) method are integrated to elucidate the mechanism underlying effect of D93 D289 protonation on binding inhibitors OV6 4B2 BACE1. The GaMD trajectory-based DL successfully identifies significant function domains. Dynamic analysis shows that strongly...

Background normalization algorithms attempt to suppress the ambient and self-noise during measurements of sonar, which enhance detection performance display effect weak signals. Conventional background methods are usually sensitive accuracy prior set filtering interval threshold, while significant noise is still detected in low frequency. In this paper, an improved algorithm proposed by thresholding processing between several local peak values valley values. Compared existing scenarios,...

Simulating the acoustic field excited by pulse sound sources holds significant practical value in computational ocean acoustics. Two primary methods exist for modeling underwater propagation time domain: Fourier synthesis technique based on frequency decomposition and time-domain model (TD-UAPM). TD-UAPMs solve wave equation domain without requiring decomposition, providing a more intuitive explanation of physical process energy over time. However, time-stepping numerical can accumulate...

Acoustic waves are essential tools for guiding underwater activities. For many years, numerical modeling of ocean acoustic propagation has been a major research focus in acoustics. Normal mode theory, one the earliest and most extensively studied methods this field, is renowned its well-established theoretical framework. The core normal theory involves solution modal equations. In classical models, these equations typically discretized using low-order finite difference methods, which, while...

Accurate calculations of ocean acoustic propagation are significant importance in various aspects such as marine environmental monitoring, oceanic communication, and navigation. Various mature simplified models have been developed, the normal mode model, parabolic approximation wavenumber integration model. The above-mentioned algorithms all limitations, namely, “applicable domain,” which restrict their range application. Instead simplifying algorithm described this article aims at solving...

It is one of the most concerning problems in hydroacoustics to find a method that can calculate acoustic propagation accurately and adapt variation seabed. Currently, one-dimensional spectral has been employed address simplified ocean model successfully. However, due model's application limitations approximation error, it poses challenges when attempting solve real-world fields. Hence, there crucial need develop direct solution for two-dimensional Helmholtz equation propagation, without...

Motivation: Integration of MRI and spatial transcriptomics may provide a new approach for imaging-genetics research probing molecular basis underlying observed phenotypes. Goal(s): We proposed pipeline integration diffusion MRI(dMRI) mouse brain to understand potential biological pathway dMRI microstructural models. Approach: spatially co-register population-averaged maps 2D transcriptomic pixelwise correlation explored genes function celltype related with metrics. Results: found FA was...

<title>Abstract</title> The oracle bone script, one of the most well-known ancient writing systems, plays a key role in study Chinese characters. To speed up digitization documents through automatic recognition, we propose recognition model called QROB (Quick Response Oracle Bone). Due to unique characteristics characters, only small portion has been translated, leading an imbalance sample distribution across different address this issue, use Inverse Image Frequency De-biasing method and...

Normal mode models are commonly used to simulate sound propagation problems in horizontally stratified oceanic environments. Although several normal have been developed, the fundamental techniques for accurately and efficiently solving modal equation still under development. Since standard three-point central finite difference scheme (SFD) has a relatively large numerical error, at least twenty sampling grid points per wavelength should be set depth direction. Herein, novel (NFD) is...

The paper presents a broadband geoacoustic inversion method based on matching the measured and modeled channel impulse responses between controlled source single hydrophone or arrays. However, motion would greatly degrade estimated response for low transmitted signals, multipath peaks be smeared. Thus wideband cross-ambiguity function is used to estimate velocity, uniform mot ion compensation Doppler scale factors conducted estimation of response. Then bottom parameters performed by...

Audio denoising methods based on deep learning have achieved great success, but these techniques usually predict the target audio instead of noise. However, in underwater acoustic signals with low SNRs, noise components dominate. To this end, a noise-aware subband attention network, namely NASA-Net, is proposed to handle task signal very SNR conditions. NASA-Net expected retain as much valuable possible while suppressing components. We design two sub-nets estimate and respectively. In...

Abstract. In the era of digital twins, high-definition 3D point clouds cultural relics, such as bronze drums ancient Southeast Asia and China, are increasingly available heritage. This study applies an automatic hierarchical clustering method to compare cluster 14 unstructured models frogs on based dissimilarity metric minimum error from 2,000 iterations global registration. Furthermore, this compares two forms presentation: surface points shape skeletons. The experimental results showed...

The distance change ratio of target can be described by characteristic feature. A method for extracting feature from LOFAR spectrum striations has been introduced. Because sound field computation and image processing consume too much time, CUDA technology used on extraction to solve the problem low efficiency. which is provided NVIDIA corporation acquired programmability perform general purpose fast running ten thousands threads concurrently. experimental results show that practical...