High Ambient Noise Research Articles

Suspected evening fish chorusing sounds detected in deep waters off Kauai, Hawaii

Noise analyses are crucial to understanding passive acoustic monitoring for marine mammals and their behavioral responses to natural and anthropogenic noises. Suspected biological sounds in the ambient noise data from hydrophones at the Pacific Missile Range Facility (PMRF), Kauai, Hawaii have been found. The sounds are suspected to be chorusing associated with the Deep Scattering Layer (DSL) upward daily vertical migration given they peak around 2 hours after sunset. They have been observed occurring daily including a period in January 2017 when high ambient noise levels during a storm was associated with baleen whale cessation of calling. The sounds occur from 1.4 to 1.8 kHz with the peak near 1650 Hz. They are detectable on multiple broadband hydrophones (21 to 68 km offshore) in deep water (1.5 to 4.7 km), suggesting a large spatial extent. The maximum spectrum level in the 1.6 kHz 1/3rd octave band observed to date was 71 dB re 1 μPa2/Hz. Individual sounds have not been identified to date given the geometries involved and the large spatial extent sensed using deep water hydrophones. These results have similarities to related chorusing research which has suggested a potential link with Myctophidae fishes.

Voice user interfaces in manufacturing logistics: a literature review

AbstractDue to the increasing digitalization in manufacturing logistics, devices to integrate the worker into the digital manufacturing system are necessary. A voice user interface (VUI) can be considered suitable for this purpose due to its flexibility and intuitive operability. Despite the popularity and acceptance of VUIs in everyday life, their use in industrial applications, especially in manufacturing logistics, is still rare. While VUIs have been successfully used in order picking for decades, hardly any other industrial fields of application exist. In this paper, we have identified various barriers to the use of VUI in industrial applications. We categorized them and identified four key barriers. We then conducted a systematic literature review to determine and compare already investigated application areas of VUIs, their characteristics, advantages and disadvantages. We found that in particular the operation of machines and industrial robots, as well as general data and information output on machine and system status, maintenance and employee training are frequently investigated. It is noticeable that VUIs are often used in combination with other user interfaces (UIs). Some challenges to VUI usage, such as high ambient noise levels, have already been solved through various approaches, while other challenges remain. Based on the results of the literature review, we put forward a research agenda regarding further suitable industrial application areas as well as general challenges for the use of VUIs in industrial environments.

Hard of hearing: the effect of distance and experimental noise on mountain chickadee song transmission

ABSTRACT Urban noise can disrupt avian communication networks, resulting in reduced communication among individuals. Mountain chickadees alter their vocalisations in areas of high ambient noise; however, it is unknown how audible their signals are within the nest cavity, where the female remains during dawn signalling. To test the relative audibility of songs within the nest box of urban and rural chickadees, we conducted a playback experiment directed at empty nest boxes using an autonomous recording unit above the box with external microphones and an internal microphone inside the nest box. We placed a speaker 25-75 m from the nest box and played two playback files of mountain chickadee songs – one with added noise and one without. Songs had a higher signal-to-noise ratio (SNR) in rural habitats and in ‘Without Added Noise’ treatments, however, SNR decreased with increasing distance from the nest box. The lower frequency song notes had a lower SNR inside the nest in both treatment types. As expected, noise and distance resulted in decreased SNR of chickadee songs within the nest cavity. These findings suggest a possible constraint in urban birds requiring them to reduce their distance from the nest to ensure audibility to the female while in noise.

Development of an Optimised Neural Network Model for RF Based UAV Detection and Identification

The rapid popularity of Unmanned Aerial Vehicles (UAVs) or drones is due to its great promise for various commercial and recreational applications. Moreover, technological advancements has made UAVs both affordable and versatile, it is able to perform task considered difficult and dangerous with high mobility, low cost and safely. Despite these advantageous applications and future potentials, UAVs are regrettably also used for malicious activities such as cyber-attack (i.e., eavesdropped the mobile phone of unsuspecting users), terrorism, drug trafficking and violation of security of restricted or sensitive areas. In order to overcome these aforementioned challenges, various UAV detection and identification techniques were developed by researchers in the past. Conventional techniques have inherent weakness such as poor performance in the presence of low visibility, less effective in an environment with high ambient noise and are also adversely affected by weather conditions. All these and more constitute major setback in UAV detection and identification. Similarly, RF sensing employed are usually based on traditional machine learning and statistical methods which are less effective in their capacity to successfully capture the rich information present in complex unstructured RF signal. To this end, a deep learning approach is proposed as a possible solution that is capable of achieving high accuracy in detection with less training time. To verify the effectiveness of our proposed convolutional neural network architecture, it was evaluated on the popular open source dataset, the DRONE RF dataset. Our empirical results demonstrates that our proposed CNN model outperformed state-of- the -art results using the same DRONE RF dataset.

Signal Processing Steady-State Surface NMR Data

Surface nuclear magnetic resonance (NMR) is a unique technique to study groundwater as it provides direct insights into water content and pore-scale properties. However, surface NMR suffers from an extremely weak signal that limits mapping speeds and can prohibit measurements in high ambient noise environments. It has recently been demonstrated that steady-state sequences can deliver orders of magnitude signal-to-noise ratio enhancement in comparison to standard methods. Most notably, the steady-state approach collapses the surface NMR signal an extremely narrow frequency band, which leads to vastly improved noise rejection. In this paper, we develop the signal processing chain for steady-state surface NMR data. Each step in the signal processing chain, transient culling, pulse windowing, de-spiking, powerline harmonics subtraction and spectral analysis are explained in detail. We demonstrate how error estimation can be efficiently performed using the power spectral density in the vicinity of the transmitting frequency. We also demonstrate how off-resonance excitation can be determined by locating the peak spectrum of surface NMR signal with an array of phase-corrected discrete Fourier transforms. The presented signal processing methods can be easily implemented and applied to extract signal features for quality control and inversion procedures. We give a field example where data collected with steady-state pulse sequences are processed and inverted.

School Hearing Screening With a Portable, Tablet-Based, Noise-Attenuating Audiometric Headset in Rural Nicaragua.

To investigate the utility and effectiveness of a noise-attenuating, tablet-based mobile health system combined with asynchronous telehealth evaluations for screening rural Nicaraguan schoolchildren for hearing loss. Prospective population-based survey. Rural Nicaraguan communities. There were 3,398 school children 7 to 9 years of age. Diagnostic automated and manual audiometry, detailed asynchronous telehealth evaluations. Referral rates, ambient noise levels, and audiometric results as well as hearing loss prevalence, types, and risk factors. Despite high ambient noise levels during screening (46.7 dBA), no effect of noise on referral rates on automated audiometry or confirmatory manual audiometry in those who failed automated testing was seen. The overall audiometric referral rate was 2.6%. Idiopathic sensorineural hearing loss (SNHL) and cerumen impaction were the most common types of hearing loss in this population with an estimated prevalence of hearing loss (all types) of 18.3 per 1,000 children. SNHL was associated with both drug exposure during pregnancy (p = 0.04) and pesticide exposure in the home (p = 0.03). Hearing screening using a tablet-based, noise-attenuating wireless headset audiometer is feasible and effective in rural low-resource environments with moderately elevated ambient noise levels. The referral rate with noise-attenuating headsets was much lower than that previous reports on this population. In addition, manual audiometry resulted in much lower referral rates than automated audiometry. The confirmed hearing loss rate in this study is comparable to reports from other low-income countries that use some form of noise attenuation during screening. Pesticide exposure and drug exposure during pregnancy are potential causes of SNHL in this population.

Using the three-station interferometry method to improve urban DAS ambient noise tomography

Distributed acoustic sensing (DAS) is a novel seismological observation technology based on the fiber-optic sensing method, and can transform existing urban fiber-optic cables into ultra-dense array for urban seismological researches, thus opening abundant opportunities for resolving fine details of near surface structures. While high frequency ambient noise recorded on DAS has been applied in surface wave tomography, it is often difficult to extract a clear dispersion curve for the data recorded by urban internet cable because of the effect of precursor signals on noise correlation functions due to uneven distribution of noise sources, and weak coupling between the cable and the solid earth. In this study, we investigate the performance of the three-station interferometry method for improving the noise cross-correlation functions of the linear array. We applied this method to a DAS dataset acquired in an urban area, suppressed the precursor signal, improved the measurement of the dispersion curve, and constructed a 2D S-wave profile that reveals the hidden fault beneath the city. We also observed that the convergence of noise cross-correlation functions with weak coupling was significantly accelerated using this method. We employed this method to improve the signal quality of surface waves at far offset for the long segment, thus obtaining a more accurate dispersion curve. In conclusion, the three-station interferometry is an effective method to enhance the surface wave signal and suppress the precursor signal retrieved from the data recorded by urban internet cable, which could help in providing high resolution images of shallow structures in built-up areas.

Semantic segmentation-based whistle extraction of Indo-Pacific Bottlenose Dolphin residing at the coast of Jeju island

Passive acoustic monitoring (PAM) is commonly utilized to monitor cetacean species’ distribution, abundance, and behavior. The demand for automated methods to detect and extract cetacean vocalizations from acoustic data has increased in the last few decades. Automatic whistle extraction of Indo-Pacific Bottlenose Dolphin (IPBD) and other whistle-producing delphinids habitating in the coastal areas represents a challenging problem due to the high ambient noise, including ship noise snapping shrimp at the same habitat. The acoustic signal containing snapping shrimp sound was usually excluded during the development of the detection method. A robust tool of bioacoustics for snapping shrimp sound is still lacking. This study trained a convolutional neural network (CNN) designed for semantic segmentation, initially developed for autonomous driving, to extract the whistle contour from spectrogram at a pixel level. A total of 1600 datasets was annotated for training and testing. As a result, the semantic segmentation classified the whistle with an overall mean Precision of 0.96, Accuracy of 0.89, and F-score of 0.86. In particular, the semantic segmentation extracted the whistle even if associated with a rich snapping shrimp sound, which the conventional method is incapable of. The advancement of metrics presented in this paper will enable long-term assessment of the IPBD population and individual or group tracking.

Acoustic Leak Localization Method for Pipelines in High-Noise Environment Using Time-Frequency Signal Segmentation

Many practical applications that involve the transportation of fluid products require the use of pipeline networks. A serious problem that emerges in this kind of networks and compromises their safety and normal operation is the occurrence of leaks. Despite the fact that extensive research has been conducted over the years, relative to the development of reliable and efficient leak detection and localization systems, not so much has been presented about pipelines in high-noise environment. This article aims to contribute to the filling of this gap. To this end, a leak localization method is proposed, based on the propagation of acoustic signals in a pipeline when a leak is present. This method employs the use of accelerometers mounted on the external surface of the monitored pipeline, in order to pick up the acoustic–vibrational leak signals. These signals are segmented both in the time and frequency domains and a time-difference-of-arrival (TDOA) algorithm along with statistical analysis is used for the identification of the leak position. The objective of this is to deal with the stochastic nature and the dispersion of the leak acoustic signals and to ensure that the proposed method can easily adapt to different pipelines and can provide efficient localization accuracy even under high-noise conditions. The proposed method was tested experimentally, both in a laboratory setup and in a refinery pipeline with high ambient noise, and the results showed that it can localize leaks efficiently with an average localization error of 4.3%.

Cognitive Distractions During Sonographic Procedures: Reducing Noise-Based and Light-Based Interruptions in Hospital Rooms

Sonographers experience a high cognitive load in hospital-based care. High ambient noise and frequent noise-based interruptions include knocking on the room door, questions from others in the room or through communication technology, alarms, alerts from personal devices, and carts and people passing in the hallway. In addition, other providers turning on the overhead light is distracting for exams that need to be conducted in reduced lighting conditions. This article suggests strategies to improve working conditions for sonographers conducting exams on a patient in the hospital room. Our strategies emerge from human factors methods and principles, which derive from communication principles and theory. These strategies are organized by reducing noise-based and light-based interruptions in the hospital room and hallway, primarily through changes to the built environment and communication technology settings and reducing the use of speech during cognitively challenging time periods through training. Most of the strategies are low-cost and can be implemented within the current built environment and communication technology infrastructure. We anticipate that these strategies could enhance patient outcomes, increase patient satisfaction, improve sonographers’ job satisfaction, protect provider health, and increase procedural efficiency.

Optimizing a rapid shutdown module for pre-polarization current in surface nuclear magnetic resonance.

Surface nuclear magnetic resonance (SNMR) technology is widely used for hydrological investigations. At high ambient noise levels, the use of a pre-polarization field can enhance the amplitude of the free induced decay (FID) signal. However, the spatial distribution of the pre-polarization field is inhomogeneous, and so it is vital to optimize the shutdown ramp and time for all combinations of the pre-polarization field amplitude and relative orientation. For this purpose, we designed a rapid and adiabatic shutdown circuit for pre-polarization surface nuclear magnetic resonance (PSNMR) systems. In the initial stage of the shutdown process, the magnetic field energy stored in the direct current transmitting coil is transferred to the capacitor by a semi-controlled full bridge circuit. At the end of the shutdown period, the pre-polarization current decays exponentially toward zero. The shutdown process can be controlled by adjusting the circuit parameters. The shutdown ramp and time of the designed circuit are verified through a series of simulations and experiments. As the initial voltage of the capacitor increases from 65 to 300V, the initial amplitude of the exponential decay of the pre-polarization current increases from 15 to 48A and the shutdown time decreases from 1.4 to 1ms. Our designscheme provides a more reliable concept for the shutdown process in the case of multiple transmitting coils. This allows a larger gain of the FID signal to be obtained in specific PSNMR applications when pre-polarization currents with slightly different channel parameters need to be shutdown uniformly.

Too Loud: Noise Exposure in Everyday Life is Causing Hearing Loss

Too Loud: Noise Exposure in Everyday Life is Causing Hearing Loss

Seismic event detection in urban Singapore using a nodal array and frequency domain array detector: earthquakes, blasts and thunderquakes

SUMMARYDetection of seismic events at or below the noise level is enabled by the use of dense arrays of receivers and corresponding advances in data analysis methods. It is not only important to detect tectonic events, but also events from man-made, non-earthquake sources and events that originate from coupling between the solid Earth and the atmosphere. In urban environments with high ambient noise levels the effectiveness of event detection methods is unclear, particularly when deployment restrictions result in an irregular receiver array geometry. Here, we deploy a dense nodal array for 1 month in the highly populated city state of Singapore. We develop a new detection method based on image processing that we call spectrogram stacking, which detects anomalous, coherent spectral energy across the array. It simultaneously detects multiple classes of signal with differing spectral content and aids event classification, so it is particularly useful for signal exploration when signal characteristics are unknown. Our approach detects more local events compared to the traditional short-term average over long-term average and waveform similarity methods, while all methods detect similar numbers of teleseismic and regional earthquakes. Local events are principally man-made non-earthquake sources, with several events from the same location exhibiting repeating waveforms. The closest earthquake occurs in peninsular Malaysia, in an area where no earthquakes have previously been detected. We also detect ground motion over a wide frequency range from discrete thunder events that show complex coupling between acoustic and elastic wavefield propagation. We suggest that care should be taken deciphering local high-frequency tectonic events in areas prone to thunder storms.

Personal audio system use can harm auditory health

Personal audio systems (PAS) are a major source of harmful noise doses, especially for children, teenagers, and young adults, who typically listen many hours daily at volumes exceeding the globally recommended public health limit of 70 dBA LEQ24h. PAS users increase listening volume to overcome high ambient noise. Research shows that (1) PAS use among people aged 9 + is associated with permanent, progressive hearing health disabilities including tinnitus, hyperacusis, speech communication impairments, and noise-induced hearing loss (NIHL); i(2) auditory health risk is highest for personal listening 1 + hours daily at over 50% volume with 5 + years of exposure; (3) clinical measures demonstrate noise-induced neural or retrocochlear auditory processing disorders precede sensory or cochlear impairments; and (4) children are highest risk because auditory system maturation is incomplete and normal hearing health is vital for learning, socialization, and vocational potential. Acquired hearing loss is associated with increased risk of accidents, communication difficulties, social isolation, and health impairments, including dementia in later life. Public health hearing conservation policies are needed to prevent an imminent NIHL epidemic in younger generations, including PAS noise emission standards and public education on the dangers of PAS use and about safer listening behaviors.

BirdNET: A deep learning solution for avian diversity monitoring

Variation in avian diversity in space and time is commonly used as a metric to assess environmental changes. Conventionally, such data were collected by expert observers, but passively collected acoustic data is rapidly emerging as an alternative survey technique. However, efficiently extracting accurate species richness data from large audio datasets has proven challenging. Recent advances in deep artificial neural networks (DNNs) have transformed the field of machine learning, frequently outperforming traditional signal processing techniques in the domain of acoustic event detection and classification. We developed a DNN, called BirdNET, capable of identifying 984 North American and European bird species by sound. Our task-specific model architecture was derived from the family of residual networks (ResNets), consisted of 157 layers with more than 27 million parameters, and was trained using extensive data pre-processing, augmentation, and mixup. We tested the model against three independent datasets: (a) 22,960 single-species recordings; (b) 286 h of fully annotated soundscape data collected by an array of autonomous recording units in a design analogous to what researchers might use to measure avian diversity in a field setting; and (c) 33,670 h of soundscape data from a single high-quality omnidirectional microphone deployed near four eBird hotspots frequented by expert birders. We found that domain-specific data augmentation is key to build models that are robust against high ambient noise levels and can cope with overlapping vocalizations. Task-specific model designs and training regimes for audio event recognition perform on-par with very complex architectures used in other domains (e.g., object detection in images). We also found that high temporal resolution of input spectrograms (short FFT window length) improves the classification performance for bird sounds. In summary, BirdNET achieved a mean average precision of 0.791 for single-species recordings, a F0.5 score of 0.414 for annotated soundscapes, and an average correlation of 0.251 with hotspot observation across 121 species and 4 years of audio data. By enabling the efficient extraction of the vocalizations of many hundreds of bird species from potentially vast amounts of audio data, BirdNET and similar tools have the potential to add tremendous value to existing and future passively collected audio datasets and may transform the field of avian ecology and conservation.

Sharing sound: Avian acoustic niches in the Brazilian Atlantic Forest

AbstractIn tropical birds, survival is dependent on the ability to effectively communicate with others against a background of high ambient noise. The Acoustic Niche Hypothesis proposes that the deterrent selective force of signal masking has caused animals sharing a habitat to partition their calling behavior among the acoustic bandwidths available, so as to minimize interference between one another. Whether and why species share the so‐called “acoustic space” remains a gap in our understanding of animal ecosystems. The aim of this study was to investigate differences between the acoustic structure of avian communities in two distinct habitats in the montane tropical forests of the Caparaó National Park in south‐eastern Brazil, and to test whether the distribution of birdsong at each conformed to the ANH. Birdsong from nine hours of passive acoustic recordings at each study site was analyzed using principal component analyses. The results contradicted the ANH, revealing strikingly similar patterns of synchronized vocal behavior (WAleixo = 0.63, p < 0.0001, RA = 0.48; WSanta Marta = 0.66, p < 0.0001, RSM = 0.41). No correlation was evident between the acoustic and temporal partitioning behavior of songbirds (pA » 0.05, dfA = 55; pSM » 0.05; dfSM = 27). This study provides a novel comparative analysis of the acoustic dynamics in two separate and diverse avian communities and support for theories of synchronized vocal behavior in such groups.Abstract in Portuguese is available with online material.

Active Sonar Target Classification with Power-Normalized Cepstral Coefficients and Convolutional Neural Network

Detection and classification of unidentified underwater targets maneuvering in complex underwater environments are critical for active sonar systems. In previous studies, many detection methods were applied to separate targets from the clutter using signals that exceed a preset threshold determined by the sonar console operator. This is because the high signal-to-noise ratio target has enough feature vector components to separate. However, in a real environment, the signal-to-noise ratio of the received target does not always exceed the threshold. Therefore, a target detection algorithm for various target signal-to-noise ratio environments is required; strong clutter energy can lead to false detection, while weak target signals reduce the probability of detection. It also uses long pulse repetition intervals for long-range detection and high ambient noise, requiring classification processing for each ping without accumulating pings. In this study, a target classification algorithm is proposed that can be applied to signals in real underwater environments above the noise level without a threshold set by the sonar console operator, and the classification performance of the algorithm is verified. The active sonar for long-range target detection has low-resolution data; thus, feature vector extraction algorithms are required. Feature vectors are extracted from the experimental data using Power-Normalized Cepstral Coefficients for target classification. Feature vectors are also extracted with Mel-Frequency Cepstral Coefficients and compared with the proposed algorithm. A convolutional neural network was employed as the classifier. In addition, the proposed algorithm is to be compared with the result of target classification using a spectrogram and convolutional neural network. Experimental data were obtained using a hull-mounted active sonar system operating on a Korean naval ship in the East Sea of South Korea and a real maneuvering underwater target. From the experimental data with 29 pings, we extracted 361 target and 3351 clutter data. It is difficult to collect real underwater target data from the real sea environment. Therefore, the number of target data was increased using the data augmentation technique. Eighty percent of the data was used for training and the rest was used for testing. Accuracy value curves and classification rate tables are presented for performance analysis and discussion. Results showed that the proposed algorithm has a higher classification rate than Mel-Frequency Cepstral Coefficients without affecting the target classification by the signal level. Additionally, the obtained results showed that target classification is possible within one ping data without any ping accumulation.

How loud is the underwater noise from operating offshore wind turbines?

Offshore wind turbines are increasingly abundant sources of underwater low frequency noise. This increase raises concern for the cumulative contribution of wind farms to the underwater soundscape and possible impact on marine ecosystems. Here, available measurements of underwater noise from different wind turbines during operation are reviewed to show that source levels are at least 10-20 dB lower than ship noise in the same frequency range. The most important factor explaining the measured sound pressure levels from wind turbines is distance to the turbines with smaller effects of wind speed and turbine size. A simple multi-turbine model demonstrates that cumulative noise levels could be elevated up to a few kilometres from a wind farm under very low ambient noise conditions. In contrast, the noise is well below ambient levels unless it is very close to the individual turbines in locations with high ambient noise from shipping or high wind speeds. The rapid increase in the number and size of offshore wind farms means that the cumulative contribution from the many turbines may be considerable and should be included in assessments for maritime spatial planning purposes as well and environmental impact assessments of individual projects.

Increased ambient noise and elevated vocal effort contribute to airborne transmission of COVID-19.

Widespread transmission of a novel coronavirus, COVID-19, has caused major public health and economic problems around the world. Significant mitigation efforts have been implemented to reduce the spread of COVID-19 but the role of ambient noise and elevated vocal effort on airborne transmission have not been widely reported. Elevated vocal effort has been shown to increase emission of potentially infectious respiratory droplets, which can remain airborne for up to several hours. Multiple confirmed clusters of COVID-19 transmission were associated with settings where elevated vocal effort is generally required for communication, often due to high ambient noise levels, including crowded bars and restaurants, meat packing facilities, and long-stay nursing homes. Clusters of COVID-19 transmission have been frequently reported in each of these settings. Therefore, analysis of COVID-19 transmission clusters in different settings should consider whether higher ambient noise levels, which are associated with increased vocal effort, may be a contributing factor in those settings. Mitigation strategies that include reduction of ambient noise, softer speech practices, and the use of technology such as microphones and speakers to decrease vocal effort will likely reduce the risk of transmitting COVID-19 or other airborne pathogens.

Acoustic, Perceptual, and Laryngoscopic Changes Post Vocal Abuse at a College Fest

College students use their voices at excessively loud intensities and long durations during college fests. The accumulative effect of vocally abusive behaviors in the presence of high ambient noise, poor vocal hygiene, and other environmental factors contribute significantly towards reducing vocal effectiveness in college students. The present study aimed to study the effect of a vocally demanding situation (college Fest) on perceptual and objective voice features of college students. In this study, a total number of 27 undergraduate female students participated in the age range of 18-27 years. Pre-test-post-test research design was applied. Data analysis consisted of video stroboscopic examination, acoustic analysis using the Multi-dimensional voice program (MDVP), and perceptual evaluation with the GRBAS scale. The data obtained were processed using descriptive statistics and non-parametric tests to determine the difference in each measured parameter pre- and postfest. Incomplete glottis closure, asymmetric vibration of vocal folds, and aperiodicity of vocal folds increased significantly postcollege Fest. GRBAS parameters also showed a change from pre Fest scores on grade (hoarseness) increased by 38%, breathiness in 34%, roughness in 26%, Asthenia in 15%, and strain increased in 38% participants. A P-value estimation revealed significantly higher scores in Jitter percent, Shimmer percent, and relative average perturbation than pre-fest scores. The results of the present study indicate a high risk of vocal disorders among college students post Fest. This study suggests a potential unmet need for college students to receive education before cultural events on optimal vocal usage, vocal hygiene, and an appropriate period of vocal rest.