In this work, a convolutional neural network (CNN) is applied to recognize acoustic spatial patterns with the aid of acoustic visualization. The acoustic spatial patterns are obtained by the singular value decomposition of an acoustic radiation operator built with the boundary integral equation. It is to explore the powerful capability of the CNN in the image processing by analogously rendering the measured acoustic spatial patterns into images. Due to practical limitations, a higher resolution of an acoustic image is achieved by interpolating the pressure on a coarse grid. Steady-state analysis of acoustic problems is a complex domain problem. The acoustic fields are then supplied into a CNN scheme as two-channel data which are real and imaginary components of the pressure. Random noises and incident waves with varying energy are added to the measured data to simulate influences from uncorrelated and correlated noises, respectively. It is demonstrated that once the CNN scheme is built and trained with adequate data, which is numerically synthesized, the patterns can be more accurately and robustly recognized by comparing it with the cross-correlation based methods. The hierarchical feature representative as well as nonlinear perception makes the proposed method a promising approach for fault diagnosis and condition monitoring based on spatial acoustic measurements.
We demonstrate a new, to the best of our knowledge, range-scalable repeatered distributed acoustic sensing (DAS) solution sharing technology with existing long-haul subsea telecom networks capable of low noise and high spatial and temporal resolution. Single ended DAS interrogation through a 2227 km line comprising 39 repeatered fiber spans is demonstrated with a 10 m spatial resolution, a 1.78 kHz sampling rate, and a strain resolution of 35 pε/√Hz for the outermost span.
Hyperspectral (HS) imaging bridges conventional imaging into spectroscopy and generates a spatial map of spectral variations. On the one hand, in HS imaging, the effect of the background on the final spectra has to be removed or managed. On the other hand, there are important classes of materials that need to be immobilized for investigation.
This article presents a spatial environmental inversion scheme using broadband impulse signals with deep learning (DL) to model a single spatially-varying sediment layer over a fixed basement. The method is applied to data from the Seabed Characterization Experiment 2022 (SBCEX22) in the New England Mud-Patch (NEMP). Signal Underwater Sound (SUS) explosive charges generated impulsive signals recorded by a distributed array of bottom-moored hydrophones.
Key Laboratory of Urban and Architectural Heritage Conservation, Ministry of Education, School of Architecture, Southeast University, 2# Sipailou, Nanjing, 210096, China.
Lightweight double leaf walls have been extensively employed in assembly and large-space buildings. Due to the complex and varied stud configurations in double leaf walls, accurately and efficiently predicting the sound transmission loss (STL) of such structures poses a significant challenge. To support performance-based design workflows, this paper presents an analytical model based on sound transmission path decoupling, enabling architects to quickly predict the STL of commonly used lightweight double leaf wall types, including wooden, steel, and acoustical stud constructions.
Dynamical theories of speech processing propose that the auditory cortex parses acoustic information in parallel at the syllabic and phonemic timescales. We developed a paradigm to independently manipulate both linguistic timescales, and acquired intracranial recordings from 11 patients who are epileptic listening to French sentences. Our results indicate that (i) syllabic and phonemic timescales are both reflected in the acoustic spectral flux; (ii) during comprehension, the auditory cortex tracks the syllabic timescale in the theta range, while neural activity in the alpha-beta range phase locks to the phonemic timescale; (iii) these neural dynamics occur simultaneously and share a joint spatial location; (iv) the spectral flux embeds two timescales-in the theta and low-beta ranges-across 17 natural languages.
