High Levels Of Background Noise Research Articles

The Importance of Sounds to Fishes

Fishes can detect underwater sounds and use them to obtain key information about the environment around them. Sounds travel rapidly over great distances in water and can provide detailed information on the presence of prey, predators, and related fishes, while the overall acoustic scene provides the fishes with key information about their environment. Although they do not have the external ears that many vertebrates have, all fish species have effective internal ears. Many fish species engage in making sounds themselves. Their calls are often produced when they are approached by other fish species, and they can be used to startle and deflect their opponents. Sounds are also produced during reproductive activities. There are often differences in the sounds made by fish species, even between closely related species. The sounds of individuals may also differ, and this may play a role in sexual selection, as males compete with one another and aim to attract females that are looking for the best males to mate with. The sounds that fishes can hear are confined to low frequencies, although this is species- dependent. It is evident that fishes can distinguish between sounds that differ in their amplitude and frequency, and also discriminate between sounds that have different temporal characteristics. They can also distinguish between sounds that arrive from different directions and distances, in some cases enabling them to locate the sources of sound. Detecting sounds may enable fishes to navigate and move to particular habitats, search for prey, move away from predators, and communicate during spawning. However, a particular problem in sound detection is the masking of those sounds that interest the fishes by high and variable levels of background noise. Although some of the background noise is generated by natural sources, including the precipitation of rain and snow, and wind and waves, many underwater sounds now come from anthropogenic sources. Some of these human-made sounds can kill or injure fishes, impair their hearing, and alter their behaviour. Interference with the detection of sounds can have especially adverse effects upon the lives of fishes. There is a need for more work on the impact of human- made underwater noise upon the fitness of fishes, and the strength of fish populations.

Lombard effect: Minke whale boing call source levels vary with natural variations in ocean noise

Minke whales were acoustically detected, localized, and tracked on the U.S. Navy's Pacific Missile Range Facility from 2012 to 2017. Animal source levels (SLs) were estimated by adding transmission loss estimates to measured received levels of 42 159 individual minke whale boings. Minke whales off Hawaii exhibited the Lombard effect in that they increased their boing call intensity in increased background noise. Minke whales also decreased the variance of the boing call SL in higher background noise levels. Although the whales partially compensated for increasing background noise, they were unable or unwilling to increase their SLs by the same amount as the background noise. As oceans become louder, this reduction in communication space could negatively impact the health of minke whale populations. The findings in this study also have important implications for acoustic animal density studies, which may use SL to estimate probability of detection.

Reciprocal Matched Filtering in the Inner Ear of the African Clawed Frog (Xenopus laevis).

Anurans (frogs and toads) are the most vocal amphibians. In most species, only males produce advertisement calls for defending territories and attracting mates. Female vocalizations are the exceptions among frogs, however in the African clawed frog (Xenopus laevis) both males and females produce distinct vocalizations. The matched filter hypothesis predicts a correspondence between peripheral auditory tuning of receivers and properties of species-specific acoustic signals, but few studies have assessed this relationship between the sexes. Measuring hearing sensitivity with a binaural recording of distortion product otoacoustic emissions, we have found that the ears of the males of this species are tuned to the dominant frequency of the female's calls, whereas the ears of the females are tuned close to the dominant frequency of the male's calls. Our findings provide support for the matched filter hypothesis extended to include male-female calling. This unique example of reciprocal matched filtering ensures that males and females communicate effectively in high levels of background noise, each sex being most sensitive to the frequencies of the other sex's calls.

A Piecewise Hybrid Stochastic Resonance Method for Early Fault Detection of Roller Bearings

An unsaturated stochastic resonance (USR) method to overcome the output saturation phenomenon observed in the classical bistable stochastic resonance (CBSR) method has been examined. However, the parameters of USR models can lead to inaccurate results while identifying the characteristic frequency amid high levels of background noise. To overcome this limitation, an adaptive piecewise hybrid stochastic resonance (APHSR) method that introduces a parameter μ to improve the performance of fault characteristic detection is proposed. The optimal parameters are determined automatically using both 3D reverse positioning and least-squares methods, combing with signal-to-noise ratio and spectral value as evaluation criteria. The significance of parameter μ is demonstrated by analyzing the critical amplitude and Kramers' escape rate. When the results were evaluated through comparison with the CBSR and USR methods via a simulation and two experiments on a motor and a parallel gearbox, it was demonstrated to be more capable of diagnosing the early faults of rotating machinery especially in high levels of background noise.

Large-scale particulate air pollution and chemical fingerprint of volcanic sulfate aerosols from the 2014–2015 Holuhraun flood lava eruption of Bárðarbunga volcano (Iceland)

Abstract. Volcanic sulfate aerosols play a key role in air quality and climate. However, the rate of oxidation of sulfur dioxide (SO2) precursor gas to sulfate aerosols (SO42-) in volcanic plumes is poorly known, especially in the troposphere. Here we determine the chemical speciation as well as the intensity and temporal persistence of the impact on air quality of sulfate aerosols from the 2014–2015 Holuhraun flood lava eruption of Icelandic volcano Bárðarbunga. To do so, we jointly analyse a set of SO2 observations from satellite (OMPS and IASI) and ground-level measurements from air quality monitoring stations together with high temporal resolution mass spectrometry measurements of an Aerosol Chemical Speciation Monitor (ACSM) performed far from the volcanic source. We explore month/year long ACSM data in France from stations in contrasting environments, close and far from industrial sulfur-rich activities. We demonstrate that volcanic sulfate aerosols exhibit a distinct chemical signature in urban/rural conditions, with NO3:SO4 mass concentration ratios lower than for non-volcanic background aerosols. These results are supported by thermodynamic simulations of aerosol composition, using the ISORROPIA II model, which show that ammonium sulfate aerosols are preferentially formed at a high concentration of sulfate, leading to a decrease in the production of particulate ammonium nitrate. Such a chemical signature is however more difficult to identify at heavily polluted industrial sites due to a high level of background noise in sulfur. Nevertheless, aged volcanic sulfates can be distinguished from freshly emitted industrial sulfates according to their contrasting degree of anion neutralization. Combining AERONET (AErosol RObotic NETwork) sunphotometric data with ACSM observations, we also show a long persistence over weeks of pollution in volcanic sulfate aerosols, while SO2 pollution disappears in a few days at most. Finally, gathering 6-month long datasets from 27 sulfur monitoring stations of the EMEP (European Monitoring and Evaluation Programme) network allows us to demonstrate a much broader large-scale European pollution, in both SO2 and SO4, associated with the Holuhraun eruption, from Scandinavia to France. While widespread SO2 anomalies, with ground-level mass concentrations far exceeding background values, almost entirely result from the volcanic source, the origin of sulfate aerosols is more complex. Using a multi-site concentration-weighted trajectory analysis, emissions from the Holuhraun eruption are shown to be one of the main sources of SO4 at all EMEP sites across Europe and can be distinguished from anthropogenic emissions from eastern Europe but also from Great Britain. A wide variability in SO2:SO4 mass concentration ratios, ranging from 0.8 to 8.0, is shown at several stations geographically dispersed at thousands of kilometres from the eruption site. Despite this apparent spatial complexity, we demonstrate that these mass oxidation ratios can be explained by a simple linear dependency on the age of the plume, with a SO2-to-SO4 oxidation rate of 0.23 h−1. Most current studies generally focus on SO2, an unambiguous and more readily measured marker of the volcanic plume. However, the long persistence of the chemical fingerprint of volcanic sulfate aerosols at continental scale, as shown for the Holuhraun eruption here, casts light on the impact of tropospheric eruptions and passive degassing activities on air quality, health, atmospheric chemistry and climate.

Comparison of speech intelligibility of elderly aged 60–69 years and young adults in the noisy and reverberant environment

In contrast to young adults with the normal hearing ability, most of the elderly can difficultly hear the speech signals in an unfavorable noisy and reverberant environment. In this paper, the effect of noise and reverberation in a room environment on the Chinese speech intelligibility of elderly (60–69 years old persons) and young adults with the normal hearing ability at different reverberation time (RT), background noise level (BNL), and signal-to-noise ratio (SNR) is investigated by using the auralization method. The results show that the speech intelligibility scores of Chinese elderly and young adults decrease with the increase in RT but increase with the increase in SNR. The obtained speech intelligibility scores of the young adults are higher than those of the elderly under the same test conditions. The speech intelligibility scores of the elderly are poorer than those of the young adults at high BNL, long RT, and low SNR. Namely, it is found that the elderly are more sensitive to the changes in reverberation case and reverberation plus noise case than the young adults. Moreover, the difference in the speech intelligibility scores between the elderly and young adults decreases with the decrease in the BNL. Lastly, it is shown that a low BNL is beneficial for improving the Chinese speech intelligibility of the elderly in a noisy and reverberant room environment.

Background noise disrupts host-parasitoid interactions.

The soundscape serves as a backdrop for acoustic signals dispatched within and among species, spanning mate attraction to parasite host detection. Elevated background sound levels from human-made and natural sources may interfere with the reception of acoustic signals and alter species interactions and whole ecological communities. We investigated whether background noise influences the ability of the obligate parasitoid Ormia ochracea to locate its host, the variable field cricket (Gryllus lineaticeps). As O. ochracea use auditory cues to locate their hosts, we hypothesized that higher background noise levels would mask or distract flies from cricket calls and result in a decreased ability to detect and navigate to hosts. We used a field manipulation where fly traps baited with playback of male cricket advertisement calls were exposed to a gradient of experimental traffic and ocean surf noise. We found that increases in noise amplitude caused a significant decline in O. ochracea caught, suggesting that background noise can influence parasitoid–host interactions and potentially benefit hosts. As human-caused sensory pollution increases globally, soundscapes may influence the evolution of tightly co-evolved host–parasitoid relationships. Future work should investigate whether female cricket phonotaxis towards males is similarly affected by noise levels.

Shallow and in depth seismic testing in urban environment: A case study in Lisbon Miocene stiff soils using joint inversion of active and passive Rayleigh wave measurements

The use of Surface Wave Methods for the characterization of the shear wave velocity profile of the soil has become increasingly attractive due to its non-invasive nature and its ability to provide information at a larger scale. Their main drawbacks are the non-uniqueness of the solution and high uncertainty of the results at higher depths. Its application in dense urban areas can be challenging due to the proximity to underground structures, high background noise level and space limitations. This paper describes a case study on the use of surface wave based methods on stiff soil in urban environment within a confined area located in Lisbon city, Portugal, surrounded by buildings and between an underground car park and the subway tunnel. The old building, built in the study area, was demolished for the construction of a new building with five basement floors. During this process, passive and active linear array and passive three-component single-station measurements were made at the surface level, before the works started, and at 2 deeper excavation levels. The results obtained using shallow and in depth measurements were compared, including HVSR and Rayleigh wave ellipticity curves and Vs profiles obtained through the conventional MASW method and through the joint inversion of both Rayleigh wave dispersion and ellipticity curves. A deeper Vs profile with low uncertainty was obtained through the joint analysis of Rayleigh wave dispersion and ellipticity curves, compatible with Vs profile obtained in depth.

Investigating the Effect of Cochlear Synaptopathy on Envelope Following Responses Using a Model of the Auditory Nerve

The healthy auditory system enables communication in challenging situations with high levels of background noise. Yet, despite normal sensitivity to pure tones, many listeners complain about having difficulties in such situations. Recent animal studies demonstrated that noise overexposure that produces temporary threshold shifts can cause the loss of auditory nerve (AN) fiber synapses (i.e., cochlear synaptopathy, CS), which appears to predominantly affect medium- and low-spontaneous rate (SR) fibers. In the present study, envelope following response (EFR) magnitude-level functions were recorded in normal hearing (NH) threshold and mildly hearing-impaired (HI) listeners with thresholds elevated above 2 kHz. EFRs were elicited by sinusoidally amplitude modulated (SAM) tones presented in quiet with a carrier frequency of 2 kHz, modulated at 93 Hz, and modulation depths of 0.85 (deep) and 0.25 (shallow). While EFR magnitude-level functions for deeply modulated tones were similar for all listeners, EFR magnitudes for shallowly modulated tones were reduced at medium stimulation levels in some NH threshold listeners and saturated in all HI listeners for the whole level range. A phenomenological model of the AN was used to investigate the extent to which hair-cell dysfunction and/or CS could explain the trends observed in the EFR data. Hair-cell dysfunction alone, including postulated elevated hearing thresholds at extended high frequencies (EHF) beyond 8 kHz, could not account for the recorded EFR data. Postulated CS led to simulations generally consistent with the recorded data, but a loss of all types of AN fibers was required within the model framework. The effects of off-frequency contributions (i.e., away from the characteristic place of the stimulus) and the differential loss of different AN fiber types on EFR magnitude-level functions were analyzed. When using SAM tones in quiet as the stimulus, model simulations suggested that (1) EFRs are dominated by the activity of high-SR fibers at all stimulus intensities, and (2) EFRs at medium-to-high stimulus levels are dominated by off-frequency contributions.

Effects of noise on mental performance and annoyance considering task difficulty level and tone components of noise.

Rotating components in mechanical systems produce tonal noises and the presence of these tones effect the quality and comfort of occupants leading to annoyance and a decrease in mental performance. The ISO 1996-2 and ANSI S1.13 standards have described metrics to quantify the effects of prominent tones, but more research on how noise attributes effect annoyance and performance, especially in different levels of task difficulty are necessary. This paper investigates relations between noise metrics, annoyance responses and mental performance under different task difficulty levels while exposed to background noise with tonal components. In this study, sixty participants were evaluated on subjective perceived annoyance and varying workloads while exposed to 18 noise signals with three different prominence tones at three frequency tones and two background noise levels while doing three different levels of n-back tasks in a controlled test chamber. Performance parameters were measured by recording the reaction time, the correct rate, and the number of misses. The results indicate an increasing trend for number of misses and reaction times at higher task difficulty levels, but a decrease for correct rate. The study results showed a significant difference for subjective responses except for annoyance and loudness under different levels of task difficulty. The participants were more annoyed with higher background noise levels, lower tone frequencies and increasing tone levels especially under increasing task difficulty. Loudness metrics highly correlate with other noise metrics. Three models for the prediction of perceived annoyance are presented based on the most strongly correlated noise metrics using neural network models. Each of the three models had different input parameters and different network structures. The accuracy and MSE of all three neural network models show it to be appropriate for predicting perceived annoyance. The results show the effect of tonal noise on annoyance and mental performance especially in different levels of task difficulty. The results also suggest that neural network models have high accuracy and efficiency, and can be used to predict noise annoyance. Model 1 is preferred in certain aspects, such as lower input parameters, making it more user-friendly. The best neural network model included both loudness metrics and tonality metrics. It seems that combined metrics have the least importance and are unnecessary in the proposed neural network model.

Effect of high levels of background noise on dog responses to a routine physical examination in a veterinary setting

Veterinary visits result in behavioural and physiological signs of fear and stress for many companion dogs. There are a number of factors that likely contribute to this response, but little is known about possible effects of the acoustic environment. The aim of this study was to assess the effect of elevated levels of common veterinary background noises on fear-related responses in dogs during a routine physical examination in a veterinary setting. Testing took place in an examination room at a veterinary clinic and involved 33 owned companion dogs. All dogs received a standardized physical examination where each dog was either presented with no additional noise (n = 16, control), or a pre-recorded noise track that included the sounds of people talking, dogs barking, and metal doors clanging (n = 17). This noise track was played back with a peak sound level of 68.0 dB, which is comparable to levels previously recorded in clinic settings. The dogs’ behavioural responses (lip licking, yawning, reduced posture, avoidance, vocalizing, trembling) were scored from video by a blinded observer for each stage of the physical examination (i.e., head exam, lymph node palpation, body palpation, temperature assessment, heart rate assessment, and respiratory rate assessment). In addition, willingness to approach the examiner was assessed before and after the examination. For behavioural measures, generalized mixed models and Fisher’s exact tests were used to assess the effects of noise, exam phase, sex, and age, with dog as a random effect. For temperature, a general linear model was used to assess the effects of noise, sex, and age, and the remaining physiological measures were assessed using t-tests. Only respiratory rate was increased with exposure to background noise (F1,31 = 6.74, p = 0.0143); no other responses were affected. However, lip licking (F5,65 = 4.04, p = 0.003), avoidance (F5,158 = 6.36, p < 0.0001), and posture reductions (F5,158 = 3.55, p = 0.0045) were increased during some exam phases. Background noise only affected a single, physiological measure during a routine exam, while exam phase seemed to have a larger influence, affecting various behavioural measures. These results suggest that ceiling effects did not prevent proper assessment of responses to noise. Thus, while noise should be minimized where possible, aspects of the examination itself should be a key focus of future research examining methods to reduce stress in dogs during veterinary examinations.

Child-directed speech in noise: Listener- and environment-related changes in speech acoustics

Talkers adapt their speech in various ways according to the demands of their listeners and the communicative context. Mothers and their preschool children participated in a real-time interactive speech production/perception paradigm, in which mothers instructed their children (or an adult listener), to select the picture corresponding to the target word. The task was performed at low and high levels of background noise (56 and 76 dB SPL, delivered through headphones), to examine the effects of decreased audibility on speech production. Acoustic-phonetic analyses of child-directed speech (CDS) and adult-directed speech (ADS) productions of target words and carrier phrase (e.g., “Find pig” ), revealed that mothers significantly enhanced the suprasegmental properties (i.e., pitch, intensity, and duration) of target words in CDS and at higher noise levels, but provided limited evidence for the hyperarticulation of the segmental properties of speech (i.e., formant frequencies of vowels, or voice-onset times of stop consonants). Results suggest that while some aspects of articulatory control are readily amenable to change as function of task/listener demands, others may not be. Understanding these capacities and constraints in the talking caregiver is relevant to theories of hyperarticulation in infant-directed speech.

Effect of trading-off office background and intermittent noise levels on performance, annoyance, distraction, and stress

In open-plan offices, work performance is affected by the acoustic environment, which includes steady-state broadband noise and intermittent, occupancy-generated noise. High levels of broadband noise (e.g., HVAC noise) can mask intermittent sounds to reduce distraction, but risk causing fatigue and other noise-related symptoms that may be detrimental to performance. In this study, the impact of the acoustic environment on work performance was investigated by adjusting the relative levels of both broadband and intermittent noise. Participants were exposed to four different acoustic environments, either starting with high background noise and low intermittent levels or vice versa. While in each background condition, the subjects performed four cognitive tasks that evaluated memory, attention, reasoning, and planning skills, respectively. Heart rate variability and electrodermal activity (EDA) were measured to gauge arousal (stress levels) in each environment. After each exposure, participants were asked to rate annoyance, distraction, fatigue, and stress, among other subjective attributes. The EDA and ratings of distraction significantly increased as the intermittent noise levels increased, while noise annoyance ratings were consistent across each background condition. Additionally, performance on the cognitive tasks was impacted by the order in which the participants experienced the acoustic environments.

Automatic sub-word unit discovery and pronunciation lexicon induction for ASR with application to under-resourced languages

We present a method enabling the unsupervised discovery of sub-word units (SWUs) and associated pronunciation lexicons for use in automatic speech recognition (ASR) systems. This includes a novel SWU discovery approach based on self-organising HMM-GMM states that are agglomeratively tied across words as well as a novel pronunciation lexicon induction approach that iteratively reduces pronunciation variation by means of model pruning. Our approach relies only on recorded speech and associated orthographic transcriptions and does not require alphabetic graphemes. We apply our methods to corpora of recorded radio broadcasts in Ugandan English, Luganda and Acholi, of which the latter two are under-resourced. The speech is conversational and contains high levels of background noise, and therefore presents a challenge to automatic lexicon induction. We demonstrate that our proposed method is able to discover lexicons that perform as well as baseline expert systems for Acholi, and close to this level for the other two languages when used to train DNN-HMM ASR systems. This demonstrates the potential of the method to enable and accelerate ASR for under-resourced languages for which a phone inventory and pronunciation lexicon are not available by eliminating the dependence on human expertise this usually requires.

DF-Fit: A Robust Algorithm for Detection of Crystallographic Information in Atom Probe Tomography Data

We report on a new algorithm for the detection of crystallographic information in three-dimensional, as retained in atom probe tomography (APT), with improved robustness and signal detection performance. The algorithm is underpinned by one-dimensional distribution functions (DFs), as per existing algorithms, but eliminates an unnecessary parameter as compared to current methods.By examining traditional DFs in an automated fashion in real space, rather than using Fourier transform approaches, we utilize an error metric based upon the expected value for a spatially random distribution for detecting crystallography. We show cases where the metric is able to successfully obtain orientation information, and show that it can function with high levels of additive and displacive background noise. We additionally compare this metric to Fourier transform methods, showing fewer artifacts when examining simulated datasets. An extension of the approach is used to aid the automatic detection of high-quality data regions within an entire dataset, albeit with a large increase in computational cost.This extension is demonstrated on acquired aluminum and tungsten APT datasets, and shown to be able to discern regions of the data which have relatively improved spatial data quality. Finally, this program has been made available for use in other laboratories undertaking their own analyses.

Listening Effort by Native and Nonnative Listeners Due to Noise, Reverberation, and Talker Foreign Accent During English Speech Perception.

Purpose Understanding speech in complex realistic acoustic environments requires effort. In everyday listening situations, speech quality is often degraded due to adverse acoustics, such as excessive background noise level (BNL) and reverberation time (RT), or talker characteristics such as foreign accent ( Mattys, Davis, Bradlow, & Scott, 2012 ). In addition to factors affecting the quality of the input acoustic signals, listeners' individual characteristics such as language abilities can also make it more difficult and effortful to understand speech. Based on the Framework for Understanding Effortful Listening ( Pichora-Fuller et al., 2016 ), factors such as adverse acoustics, talker accent, and listener language abilities can all contribute to increasing listening effort. In this study, using both a dual-task paradigm and a self-report questionnaire, we seek to understand listening effort in a wide range of realistic classroom acoustic conditions as well as varying talker accent and listener English proficiency. Method One hundred fifteen native and nonnative adult listeners with normal hearing were tested in a dual task of speech comprehension and adaptive pursuit rotor (APR) under 15 acoustic conditions from combinations of BNLs and RTs. Listeners provided responses on the NASA Task Load Index (TLX) questionnaire immediately after completing the dual task under each acoustic condition. The NASA TLX surveyed 6 dimensions of perceived listening effort: mental demand, physical demand, temporal demand, effort, frustration, and perceived performance. Fifty-six listeners were tested with speech produced by native American English talkers; the other 59 listeners, with speech from native Mandarin Chinese talkers. Based on their 1st language learned during childhood, 3 groups of listeners were recruited: listeners who were native English speakers, native Mandarin Chinese speakers, and native speakers of other languages (e.g., Hindu, Korean, and Portuguese). Results Listening effort was measured objectively through the APR task performance and subjectively using the NASA TLX questionnaire. Performance on the APR task did not vary with changing acoustic conditions, but it did suggest increased listening effort for native listeners of other languages compared to the 2 other listener groups. From the NASA TLX, listeners reported feeling more frustrated and less successful in understanding Chinese-accented speech. Nonnative listeners reported more listening effort (i.e., physical demand, temporal demand, and effort) than native listeners in speech comprehension under adverse acoustics. When listeners' English proficiency was controlled, higher BNL was strongly related to a decrease in perceived performance, whereas such relationship with RT was much weaker. Nonnative listeners who shared the foreign talkers' accent reported no change in listening effort, whereas other listeners reported more difficulty in understanding the accented speech. Conclusions Adverse acoustics required more effortful listening as measured subjectively with a self-report NASA TLX. This subjective scale was more sensitive than a dual task that involved speech comprehension, which was beyond sentence recall. It was better at capturing the negative impacts on listening effort from acoustic factors (i.e., both BNL and RT), talker accent, and listener language abilities.

Environmental noise influences song frequency of Yellow Warblers (Dendroica petechia) in Grand Teton National Park

We explored how Yellow Warblers (Dendroica petechia) alter their songs when encountering noise in Grand Teton National Park. Different strategies for avoiding signal masking are used by other species of birds, yet there is a lack of information of birds’ responses to higher noise levels -- above 65 dB; such levels are often found in National Parks that have many visitors. In this study, we investigated singing behavior of Yellow Warblers when facing noise that ranged from 30 dB to 80 dB. In these preliminary results, we found that some features of Yellow Warblers did not appear to change with background noise level, including mean minimum frequency, bandwidth and song length. Other song features we studied did show small but statistically significant changes with higher background noise, including the peak frequency and the mean minimum frequency, both of which were significantly negatively correlated with the level of background noise. This result is different from the positive correlations that are typically observed. We speculate that this difference is due to the very high dB levels of background noise that we observed.&#x0D; &#x0D; Featured photo by wagon16 on Flickr. https://flic.kr/p/G2W6Bk

Feature Extraction of Visual Evoked Potentials Using Wavelet Transform and Singular Value Decomposition

Introduction: Brain visual evoked potential (VEP) signals are commonly known to be accompanied by high levels of background noise typically from the spontaneous background brain activity of electroencephalography (EEG) signals. Material and Methods: A model based on dyadic filter bank, discrete wavelet transform (DWT), and singular value decomposition (SVD) was developed to analyze the raw data of visual evoked potentials and extract time-locked signals with external visual stimulation. A bio-amplifier (iERG 100P) and data acquisition system (OMB-DAQ-3000) were utilized to record EEG raw data from the human scalp. MATLAB Data Acquisition Toolbox, Wavelet Toolbox, and Simulink model were employed to analyze EEG signals and extract VEP responses. Results: Results were verified in Simulink environment for the real recorded EEG data. The proposed model allowed precise decomposition and classification of VEP signals through the combined operation of DWT and SVD. DWT was successfully used for the decomposition of VEP signals to different frequencies followed by SVD for feature extraction and classification. Conclusion: The visual and quantitative results indicated that the impact of the proposed technique of combining DWT and SVD was promising. Combining the two techniques led to a two-fold increase in the impact of peak signal to noise ratio of all the tested signals compared to using each technique individually.

Learning deep features of spectrogram to estimate acoustic density of a fish species inhabitating Posidonia oceanica seagrasses

In field acoustic investigations on seagrass ecology, characteristic /kwa/ sound production of a fish species systematically present in meadows constituted by Posidonia oceanica species were recorded (Mediterranean, since 1995). Machine learning is investigated to automate estimation of acoustic density of the /kwa/ species. A multi-model deep neural network is implemented that directly inputs the spectrograms of signals recorded on hydrophones above and within the canopy. The network is a combination of 1D convolutional neural networks and a recurrent neural network with long short-term memory units and fully connected networks. Such architecture is able to process spectral information in the convolutional network and temporal information in the recurrent network, thereby making the best use of information in both time and frequency domains. In spite of high background noise level, the true-positive detection rate is shown to be excellent. With respect to human expert supervision, false-positive detections were, for the major part, faint sounds produced by distant fishes of the same species, as verified through careful re-listening. The apparent error rate is 8.4 ×10−2 on a partially mislabeled dataset of three days (7-10 pm). The true error rate, after label correction, is estimated at 1×10−2.

Individual Classification of Single Trial EEG Traces to Discriminate Brain responses to Speech with Different Signal-to-Noise Ratios.

To gain knowledge of listening effort in adverse situations, it is important to know how the brain processes speech with different signal-to-noise ratios (SNR). To investigate this, we conducted a study with 33 hearing impaired individuals, whose electroencephalographic (EEG) signals were recorded while listening to sentences presented in high and low levels of background noise. To discriminate between these two conditions, features from the 64-channel EEG recordings were extracted using the power spectrum obtained by a Fast Fourier Transform. Features vectors were selected on an individual basis by using the statistical R2 approach. The selected features were then classified by a Support Vector Machine with a nonlinear kernel, and the classification results were validated using a leave-one-out strategy, and presented an average classification accuracy over all 33 subjects of 83% (SD=6.4%). The most discriminative features were selected in the high-beta (19-30 Hz) and gamma (30-45 Hz) bands. These results suggest that specific brain oscillations are involved in addressing background noise during speech stimuli, which may reflect differences in cognitive load between the conditions of low and high background noise.