Construction noise is one of the main sources of noise pollution in many cities, and degrades the comfort level of living spaces. It was previously reported that a noise barrier with a wide “cross-sectional profile” (e.g., T- or Y-shaped) could enhance the noise attenuation performance, and the jagged edge “longitudinal profile” on the top edge of the noise barrier could generate destructive interference sound fields behind the noise barrier, which could further reduce the noise levels. The present paper attempts to study the noise attenuation performances of jagged edge profiles applied on the edge of a cantilever, which was mounted at the top of a commercial passive noise barrier. In addition to the numerical simulations, the full-sized prototypes were also experimentally tested on a construction site with noise generated by a boring machine. Both numerical simulation and experimental results showed that this barrier with slanted flat-tip jagged cantilever would perform better than the traditional barrier having a Straight edge cantilever of same height, with a maximum additional attenuation of 5.0 dBA obtained experimentally. The barrier with a slanted flat-tip jagged cantilever could also extend the shadow zone behind the barrier to higher levels.

The aim of this paper is to provide an overview of the existing practices used for acoustic signal processing of noise of machines in various industries. There has been a surge in deep learning based methods for acoustic detection and classification of machinery fault diagnosis. This paper reviews the deep learning models, including the convolutional neural networks, the recurrent neural networks, the spiking neural networks, among the other variants of neural network models specific to industrial noise. Important applications such as sound detection, localization, directivity, source separation are discussed, to aid condition monitoring.

No abstract

Automated techniques to detect Alzheimer’s Dementia through the use of audio recordings of spontaneous speech are now available with varying degrees of reliability. Here, we present a systematic comparison across different modalities, granularities and machine learning models to guide in choosing the most effective tools. Specifically, we present a multi-modal approach (audio and text) for the automatic detection of Alzheimer’s Dementia from recordings of spontaneous speech. Sixteen features, including four feature extraction methods (Energy–Time plots, Keg of Text Analytics, Keg of Text Analytics-Extended and Speech to Silence ratio) not previously applied in this context were tested to determine their relative performance. These features encompass two modalities (audio vs. text) at two resolution scales (frame-level vs. file-level). We compared the accuracy resulting from these features and found that text-based classification outperformed audio-based classification with the best performance attaining 88.7%, surpassing other reports to-date relying on the same dataset. For text-based classification in particular, the best file-level feature performed 9.8% better than the frame-level feature. However, when comparing audio-based classification, the best frame-level feature performed 1.4% better than the best file-level feature. This multi-modal multi-model comparison at high- and low-resolution offers insights into which approach is most efficacious, depending on the sampling context. Such a comparison of the accuracy of Alzheimer’s Dementia classification using both frame-level and file-level granularities on audio and text modalities of different machine learning models on the same dataset has not been previously addressed. We also demonstrate that the subject’s speech captured in short time frames and their dynamics may contain enough inherent information to indicate the presence of dementia. Overall, such a systematic analysis facilitates the identification of Alzheimer’s Dementia quickly and non-invasively, potentially leading to more timely interventions and improved patient outcomes.

No abstract