Deep learning for unsupervised separation of environmental noise sources

Unattended and continuously running environmental noise monitoring systems can capture an intractable amount of data. The signals captured can include a multitude of sources (e.g., wind noise and anthropogenic noise sources) in addition to the environmental noise sources of interest (e.g., aircraft, vehicles, trains, and military weapons). In this presentation, we explore the use of machine-learning methods to effectively isolate and identify environmental noise sources captured on such a noise monitoring system. Specifically, we consider the use of both unsupervised (e.g., principle components analysis, clustering methods, and deep belief networks) and supervised (e.g., logistic regression, support vector machines, and neural networks) pattern-learning methods to derive the features of interest and classify the signals based on the obtained features. The generalization performance of each method is assessed using a dataset of over 120,000 human classified signals, and the strengths and weaknesses of each approach are discussed.

Environmental noise monitoring systems continuously measure sound levels without assigning these measurements to different noise sources in the acoustic scenes, therefore incapable of identifying the main noise source. In this paper a feasibility study is presented on a new monitoring concept in which an acoustic pattern classification algorithm running in a wireless sensor is used to automatically assign the measured sound level to different noise sources. A supervised noise source classifier is learned from a small amount of manually annotated recordings and the learned classifier is used to automatically detect the activity of target noise source in the presence of interfering noise sources. The sensor is based on an inexpensive credit-card-sized single-board computer with a microphone and associated electronics and wireless connectivity. The measurement results and the noise source information are transferred from the sensors scattered around the measurement site to a cloud service and a noise portal is used to visualise the measurements to users. The proposed noise monitoring concept was piloted on a rock crushing site. The system ran reliably over 50days on site, during which it was able to recognise more than 90% of the noise sources correctly. The pilot study shows that the proposed noise monitoring system can reduce the amount of required human validation of the sound level measurements when the target noise source is clearly defined.

Reducing environmental noise in urban settings, i.e., unwanted or harmful outdoor sounds produced by human activity, has become an important issue in recent years. Most countries have established regulations that set maximum permitted noise levels. However, enforcing these regulations effectively remains challenging as it requires active monitoring networks and audio analysis performed by trained specialists. The manual evaluation of the audio recordings is laborious, time-consuming, and inefficient since many audios exceeding the noise level threshold do not correspond to a sound event considered by the regulation. To address this challenge, this work proposes a computational pipeline to assist specialists in detecting noise sources in the built environment that do not comply with the Chilean noise regulation. The system incorporates a deep neural model following a pre-trained audio neural network architecture transferred to a dataset compiled from public sources and recordings in Valdivia, Chile. The target dataset follows a customized taxonomy of urban sound events. The system also uses a public API so potential users can post audio files to obtain a prediction matrix reporting the presence of noise sources contributing to environmental noise pollution. Experiments using recordings from two continuous noise monitoring stations showed that the amount of data to be inspected by the specialist is decreased by 97% when the deep-learning tools are used. Therefore, this system efficiently assists trained experts in enforcing noise legislation through machine-assisted environmental noise monitoring.

Environmental noise pollution is a significant concern for public health and well-being. It is important to accurately identify and classify environmental noise sources to develop effective noise reduction strategies. This paper examines using an audio spectrogram transformer (AST) for environmental noise tagging tasks. The AST is a pure attention-based model that takes the spectrograms of the audio signals as input and calculates the self-attention without convolutions. Previously, it was pre-trained on large datasets such as ImageNet and AudioSet, showing higher precision than prior work. Although the hyperparameters were given, many have not been clear from the previous literature. Results show that there are a few choices for the patch split overlap, more overlap does not result in significantly improved performance. It is also shown that instead of the default 128 frequency bins, 96 is another choice, which can reduce the computations. The results further show that 30% - 40% can be masked for the frequency, 20% - 50% for the time dimension. The trained model is further tested on the dataset of different environmental noise sources collected by SiteHive Hexanodes across Australia and New Zealand. Results show that the AST model can achieve high accuracy in identifying different environmental noises.

: The airborne microwave remote sensing was performed during EuroSTARRS using the Naval Research Laboratory STARRS instrument (Salinity, Temperature, and Roughness Remote Scanner). This leading edge passive L-Band instrument was acquired by NRL shortly before the campaign. Development of techniques for its use, and studies to understand its performance, are important parts of the acquisition process. In this paper we will examine instrument calibration and noise issues for STARRS which are associated with the flights conducted as part of the EuroSTARRS campaign, from November 16-23, 2001 in Germany, France and Spain. The issues we will address here are: pitch and roll improvements, L-Band absolute and ocean calibration, instrument stability, and environmental noise sources. A strong environmental noise source, located in the direction of Barcelona, was noticed during the Casablanca flight. A second environmental source was noted in the analysis later. The high altitude section of the Casablanca flight has elevated TB (most notably in all the outboard beams) relative to the low altitude section of the flight. The effects such environmental (presumably man-made) sources. Will have on satellite observations of L-Band brightness temperature may be significant.

Environmental noise monitoring is an important link in improving the quality of life and strengthening environmental protection. In the bustling neighborhoods and residential areas of major cities, large-scale of environmental noise pollution have erected the streets. The design is mainly based on MCS89C5 1 microcontroller, which is the core of the audio detection system design. Capacitive electret microphone work by the principle of the sensor, the sound signal converts into electrical signal output when the microphone receiving noise, and then through the hardware circuit: amplify the signal, filtering A / D conversion circuit and software program control to complete the sound frequency detection function. Detection of the sound frequency display through the LED digital display, this design of the audio detection system is low cost, portable, can display environment noise digitally and so on. This can be widely used in industrial and mining enterprises, institutions and schools that need to measure and control the environment noise.

BackgroundThe urban soundscape, which represents the totality of noise in the urban setting, is formed from a wide range of sources. One of the most ubiquitous and least studied of these is street-level (i.e., sidewalk) noise. Mainly associated with vehicular traffic, street level noise is hard to ignore and hard to escape. It is also potentially dangerous, as excessive noise from any source is an important risk factor for adverse health effects. This study was conducted to better characterize the urban soundscape and the role of street level noise on overall personal noise exposure in an urban setting.MethodsStreet-level noise measures were obtained at 99 street sites located throughout New York City (NYC), along with data on time, location, and sources of environmental noise. The relationship between street-level noise measures and potential predictors of noise was analyzed using linear and logistic regression models, and geospatial modeling was used to evaluate spatial trends in noise. Daily durations of street-level activities (time spent standing, sitting, walking and running on streets) were estimated via survey from a sample of NYC community members recruited at NYC street fairs. Street-level noise measurements were then combined with daily exposure durations for each member of the sample to estimate exposure to street noise, as well as exposure to other sources of noise.ResultsThe mean street noise level was 73.4 dBA, with substantial spatial variation (range 55.8-95.0 dBA). Density of vehicular (road) traffic was significantly associated with excessive street level noise levels. Exposure duration data for street-level noise and other common sources of noise were collected from 1894 NYC community members. Based on individual street-level exposure estimates, and in consideration of all other sources of noise exposure in an urban population, we estimated that street noise exposure contributes approximately 4% to an average individual’s annual noise dose.ConclusionsStreet-level noise exposure is a potentially important source of overall noise exposure, and the reduction of environmental sources of excessive street- level noise should be a priority for public health and urban planning.Electronic supplementary materialThe online version of this article (doi:10.1186/s12940-015-0006-y) contains supplementary material, which is available to authorized users.

Governments and commissions all around the globe have worked on developing policies and rules as well as actions and measures for continuous assessment and remedies for environmental noise pollution. All of which requires continuous and accurate measurements of noise levels. The advent of smart city and IoT-based monitoring makes it possible for different technologies to cooperate in collecting and reporting environmental noise levels for longer duration and wider geographical region. Among others, mobile crowdsourcing (MCS) appears to be a promising technique for environmental noise monitoring with minimal upfront cost and almost no recurrent cost for regulators. This is due to the assumption of voluntary participation from mobile device owners who may be reluctant to participate if the recurrent cost of mobile resource consumption is high. In this work, an approach to improve mobile device resources efficiency for mobile crowdsourcing application in environmental noise monitoring is proposed. That is, noise samples from a device are collected only if it is significant to the accuracy of the measurements. Also this paper defines and mathematically formalizes the problem under study and develops two algorithms to optimize both resource and measurements efficiency. The results demonstrate the efficiency of the proposed solution.

Climate change, noise, water and air pollution, as well as other environmental problems, are becoming increasingly serious due to the rapid growth of the human population, industrialization, infrastructure improvements, the development of cars, and the use of fossil fuels. It is crucial to keep an eye on these problems and offer strategies to solve them in order to guarantee a healthy life and a brighter future. A new area of study that can help with the monitoring of noise and air pollution is called "smart sensor networks," which combines computer science, wireless communication, and electronics. Here, we propose an IoT-based system for monitoring air quality and noise pollution that will enable us to keep an eye on and assess these factors in specific locations. The system monitors the air for the presence of dangerous gases or compounds using a variety of wind sensors, and it continuously updates this information on a compact controller. Additionally, our system keeps track of audio intensity and transmits this information to an online server using IoT. The microcontroller that interprets this data and sends it over the Internet communicates with the sensors. This enables the government to keep an eye on, manage, and take action against air pollution in diverse locations. Additionally, respected authorities may keep an eye on noise pollution around hospitals, schools, and other non-wooded places. If the system notices problems with air quality or noise, it will alert authorities so they can take appropriate measures to mitigate the problem.

A study was conducted to measure background noise levels and to predict reverberation times in university classrooms. Twenty-five classrooms were selected due to the dimensions and location. Background noise level in all classrooms exceeds the recommendations in ANSI S12.60-2002. Calculated reverberation times based on the Sabine equation indicate that none of the classrooms have reverberation times below the recommendations in ANSI S12.60-2002, as well as the Building Bulletin 93. It is concluded that interior noise sources, such as classrooms’ HVACs, students from other classrooms or corridors, and corridors’ HVACs, must be taken into consideration rather than exterior noise sources because all eight buildings were located far from any environmental noise sources. The classrooms’ doors (due to the openings) are more responsible for sound transmission than are the walls.

Currently, noise pollution deriving from traffic, industry and leisure activities is one of the main environmental problems that affect people quality of life and health. However, it is very difficult to quantify noise effects, because they depend on people noise level tolerance, environmental noise sources, methods of definition of noise exposure and indicators. The acoustic descriptors currently used, indeed, are hardly understandable for those who do not deal with the topic, thus limiting the comprehension of the damage caused by noise. Consequently, it is essential to provide information that is easier to understand and closer to the noise pollution perceived by people. Innovative acoustic indices, like the Harmonica Index, have been defined in recent years, to overcome the aforementioned limits. Their application to different spatial contexts, with the identification of the background and the event-based components of the environmental noise, can contribute to understand the influence of human activities on the quality of an acoustic environment. In this paper, starting from data acquired both in crowdsourcing and with traditional instruments, the application of the Harmonica Index to urban, peri-urban and rural areas in south Italy will show their acoustic characteristics that could be understood also by a non-technical public.

BackgroundWhile the adverse health effects of civil aircraft noise are relatively well studied, impacts associated with more intense and intermittent noise from military aviation have been rarely assessed. In recent years, increased training at Naval Air Station Whidbey Island, USA has raised concerns regarding the public health and well-being implications of noise from military aviation.ObjectiveThis study assessed the public health risks of military aircraft noise by developing a systematic workflow that uses acoustic and aircraft operations data to map noise exposure and predict health outcomes at the population scale.MethodsAcoustic data encompassing seven years of monitoring efforts were integrated with flight operations data for 2020–2021 and a Department of Defense noise simulation model to characterize the noise regime. The model produced contours for day-night, nighttime, and 24-h average levels, which were validated by field monitoring and mapped to yield the estimated noise burden. Established thresholds and exposure-response relationships were used to predict the population subject to potential noise-related health effects, including annoyance, sleep disturbance, hearing impairment, and delays in childhood learning.ResultsOver 74,000 people within the area of aircraft noise exposure were at risk of adverse health effects. Of those exposed, substantial numbers were estimated to be highly annoyed and highly sleep disturbed, and several schools were exposed to levels that place them at risk of delay in childhood learning. Noise in some areas exceeded thresholds established by federal regulations for public health, residential land use and noise mitigation action, as well as the ranges of established exposure-response relationships.Impact statementThis study quantified the extensive spatial scale and population health burden of noise from military aviation. We employed a novel GIS-based workflow for relating mapped distributions of aircraft noise exposure to a suite of public health outcomes by integrating acoustic monitoring and simulation data with a dasymetric population density map. This approach enables the evaluation of population health impacts due to past, current, and future proposed military operations. Moreover, it can be modified for application to other environmental noise sources and offers an improved open-source tool to assess the population health implications of environmental noise exposure, inform at-risk communities, and guide efforts in noise mitigation and policy governing noise legislation, urban planning, and land use.

Multichannel feedforward active noise control system combined with noise source separation by microphone arrays

Environmental noise, as well as being a form of environmental pollution that affects mainly urban areas, it is a problem that involves people's quality of life. This paper presents a methodology that takes advantage of the Volunteered Geographic Information (VGI) in obtaining georeferenced environmental noise maps and the corresponding statistics for a particular area of interest. The methodology herein presented considers the phases of Data acquisition, Analysis and data processing, and Visualization of the information. The design and development of a Geographic Information System is presented, which consists of a web mapping application, an application for mobile devices, called NoiseMonitor, Geospatial Analysis and Machine Learning methods (Support Vector Machines and Artificial Neural Networks) for acoustic noise prediction using contextual information, that is, some factors around the measurements. A study case is based on Mexico City.

Background:Noise is recognized as a source of hazard to the patient's environment. Studies have also shown that it has a direct impact on mortality and morbidity as a result of sleep deprivation which affects the immunity of critically ill patients.Objectives:The aim of this study was to measure levels of environmental noise in a six- bed, open-plan general medical intensive care unit (MICU).Methods:Levels of exposure to environmental noise were assessed in the intensive care unit of King Fahad Hospital of the University (KFHU) where measurements of environmental noise were taken using calibrated sound level meter during shifts of working days and weekends.Results:Statistical analysis revealed that there were no significant differences between noise levels in the morning, evening and night shifts of working days and weekends in the ICU of KFHU (p value =0.155, 0.53 and 0.711) respectively. There was no significant difference between overall level of exposure to noise in the working days and weekends as well (p-value=0.71). However, the assessed levels of exposures to noise were still higher than stipulated international standards.Conclusion:Some sources of environmental noise, such as the use of oxygen, suction equipment or respirators are unavoidable. Nevertheless, hospital ICUs should have measures to minimize the level of exposure to noise in the ICU.Further research in this area might focus on the noise level and other modifiable environmental stress factors in the ICU that affect patients as well as the staff.