Auditory Model Research Articles

LIAeq100ms, An A-duration adjusted impulsive noise damage risk criterion

Impulsive noise damage risk criteria (DRCs) have been the subject of much debate nationally and internationally for more than 30 years. Several approaches have been used in proposed DRCs including: curves defining exposure based on peak level and A or B duration; auditory hazard units based on the analytical auditory model known as AHAAH; and LAeq metrics based on the equal energy concept. Each of the approaches has positive and negative attributes. One of the issues with LAeq metrics has been the over estimation of hazard with long duration impulses such as those coming from blasts, artillery, large mortars, or shoulder launched missiles. The presentation will describe and discuss the LIAeq100ms impulsive DRC which includes an adjustment based on the A-duration of the impulsive noise and a method of computing protected impulsive noise exposures using impulsive insertion loss data from hearing protectors measured with methods defined in ANSI S12.42.

Modeling auditory coding: from sound to spikes.

Models are valuable tools to assess how deeply we understand complex systems: only if we are able to replicate the output of a system based on the function of its subcomponents can we assume that we have probably grasped its principles of operation. On the other hand, discrepancies between model results and measurements reveal gaps in our current knowledge, which can in turn be targeted by matched experiments. Models of the auditory periphery have improved greatly during the last decades, and account for many phenomena observed in experiments. While the cochlea is only partly accessible in experiments, models can extrapolate its behavior without gap from base to apex and with arbitrary input signals. With models we can for example evaluate speech coding with large speech databases, which is not possible experimentally, and models have been tuned to replicate features of the human hearing organ, for which practically no invasive electrophysiological measurements are available. Auditory models have become instrumental in evaluating models of neuronal sound processing in the auditory brainstem and even at higher levels, where they are used to provide realistic input, and finally, models can be used to illustrate how such a complicated system as the inner ear works by visualizing its responses. The big advantage there is that intermediate steps in various domains (mechanical, electrical, and chemical) are available, such that a consistent picture of the evolvement of its output can be drawn. However, it must be kept in mind that no model is able to replicate all physiological characteristics (yet) and therefore it is critical to choose the most appropriate model—or models—for every research question. To facilitate this task, this paper not only reviews three recent auditory models, it also introduces a framework that allows researchers to easily switch between models. It also provides uniform evaluation and visualization scripts, which allow for direct comparisons between models.

Acoustic feature extraction method for robust speaker identification

When there is a mismatch between the acoustic training environment and the testing environment, the performance of automatic speaker identification systems degrades significantly. A robust feature extraction method for speaker recognition based on the gammatone filter is therefore proposed in this paper. By employing the working mechanism of the human auditory model instead of the traditional triangular filter banks, gammatone filter banks are used to simulate the auditory model of the human ear cochlea. The cube root compression method, equal loudness technology, and relative spectral (RASTA) filtering technology are incorporated into the robust feature extraction process. A simulation experiment is conducted based on the Gaussian mixture model (GMM) recognition algorithm. The experimental results indicate that the proposed feature parameters could show superior robustness and represent the characteristics of the speaker better than the conventional mel-frequency cepstrum coefficient (MFCC), cochlear cepstrum coefficient (CFCC) and relative spectra-perceptual linear predictive (RASTA-PLP) parameters.

Speech Practice Effects on Bilingual Children's Fast Mapping Performance.

Learning a new word involves many subsystems and their interactions. The purpose of this study was to examine whether speech practice facilitates the subsequent fast mapping performance of bilingual preschool children. Participants were 18 typically developing preschoolers who learned Cantonese (L1) as a home language and English (L2) as a second language. Each participant was asked to repeat each of the four novel words after an auditory model (speech practice) before he or she was exposed to the novel word-objects in the subsequent fast mapping task. Significant speech practice and language effects on fast mapping production scores were found. These findings suggest a complex relationship between speech learning, fast mapping, and L1-L2 language skills. Our results suggest that clinical methods that facilitate bilingual children's production skills (e.g., repeat after a model) hold promise as a means of improving the initial stage of word learning in both languages.

Tuning the cognitive environment: Sound masking with “natural” sounds in open-plan offices

With the gain in popularity of open-plan office design and the engineering efforts to achieve acoustical comfort for building occupants, a majority of workers still report dissatisfaction in their workplace environment. Office acoustics influence organizational effectiveness, efficiency, and satisfaction through meeting appropriate requirements for speech privacy and ambient sound levels. Implementing a sound masking system is one tried-and-true method of achieving privacy goals. Although each sound masking system is tuned for its specific environment, the signal—random steady state electronic noise, has remained the same for decades. This session explores how “natural” sounds may be used as an alternative to this standard masking signal employed so ubiquitously in sound masking systems in the contemporary office environment. As an unobtrusive background sound, possessing the appropriate spectral characteristics, this proposed use of “natural” sounds for masking challenges the convention that masking sounds should be as meaningless as possible. Based on psychophysical data and a sound-field analysis through an auditory model, we hypothesize that “natural” sounds as masking sounds have the ability (with equal success as conventional masking sounds) to meet standards and criteria for speech privacy while enhancing cognitive functioning, optimizing the ability to concentrate, and increasing overall worker satisfaction.

Envelope-power based prediction of auditory masking and speech intelligibility

In natural surroundings, people are often part of challenging acoustical scenarios (noise) interferers and reverberation. Amplitude modulations (AM) are fundamental acoustic features used to extract information in such situations. It has been shown that the envelope power spectrum model [EPSM; Ewert and Dau, J. Acoust. Soc. Am. 108, 1181 (2000)] can account for psychoacoustic AM detection and masking data by considering the long-term envelope signal-to-noise-ratios (SNR) at the output of modulation filters. Recently, this concept was extended to consider short-term envelope SNRs to predict speech intelligibility in fluctuating maskers and in reverberation [mr-sEPSM; Jørgensen et al., J. Acoust. Soc. Am. 134, 1475 (2013)]. In this study, the EPSM was further extended to consider envelope power SNRs (AM cues) on various time scales and “classic” power SNRs (intensity cues). The goal was to develop an auditory model for jointly predicting psychoacoustic masking and speech intelligibility. The model was demonstrated to account for a broad variety of psychoacoustic effects as AM detection, AM discrimination, AM masking, spectral masking, forward masking, intensity just-noticeable-differences (JNDs) and hearing threshold. Speech intelligibility was predicted comparable to Jørgensen et al.. Implications for future modeling including applications to audio and speech quality assessment are discussed.

Conventional Acoustic Features Based Gammachirp Filterbank for Text Independent Speaker Identification System in Noisy Environments

In this paper we study the performance of a novel feature extraction for both of MFCC and PLP features based Gammachirp filterbank in a state-of-the-art text independent speaker identification system (SID). The novel feature for extracting these parameters is based on an auditory periphery model for robust speaker identification system which mimics the human auditory system characteristics and relies on the Gammachirp filterbank to imitate the cochlea frequency resolution with nonlinear resolution according to the equivalent rectangular bandwidth (ERB) scale. Our evaluations show that the proposed feature performs considerably better than conventional acoustic features. We further demonstrate that integrating the proposed feature in various noisy environments yields promising recognition performance.

“Total inner memory”: Deliberate uses of multimodal musical imagery during performance preparation.

Expert pianists are expected to memorize a large and complex repertoire, but there is a lack of clear memorization pedagogy and considerable variation in the amount and type of practice undertaken by experts (Jorgensen, 2002). Deliberate imagery has been advocated as a potential means of enhancing memorization and performance quality (Gieseking & Leimer, 1932; Gregg, Clark, & Hall, 2008; Holmes, 2005), improving practice efficiency and reducing the amount of physical required for secure performance (Connolly & Williamon, 2004; Freymuth, 1999; Gieseking & Leimer, 1932). Some experimental evidence supports these views, although precise definitions of imagery rehearsal, and methods for measuring its effects, have varied. In general, a number of studies have found that imagery plus physical is equal to physical alone (Bernardi, Schories, Jabusch, Colombo, & Altenmuller, 2013; Coffman, 1990; Ross, 1985; Theiler & Lippman, 1995), while others have found that imagery plus physical is superior to physical alone (Rubin-Rabson, 1937, 1941). The term mental imagery rehearsal is adopted here, in preference to others such as practice or rehearsal, to express a vivid sense of the activity as an imaginative and constructive act. It also aims to better encapsulate the notion that imagery occurs during performance and that and physical processes cannot be entirely separated.Deliberate imagery techniques are accepted to be of value by many musicians, and have been shown by some studies to enhance learning (Driskell, Copper, & Moran, 1994; Rubin- Rabson, 1937). However, instrumental music pedagogy does not appear to incorporate their use widely (Holmes, 2005). Learning and memorization skills are rarely taught explicitly (Ginsborg, 2004), and skills training programs are not widely applied within the performing arts (Clark & Williamon, 2011; Hays, 2002). While several types of for music performance have been identified (Connolly & Williamon, 2004; Holmes, 2005), and a small number of detailed studies have investigated some of the ways in which musicians implement imagery techniques in practice (Bailes, 2009; Holmes, 2005), it is not yet clear how or to what extent the teaching and implementation of imagery techniques may benefit musicians. There is still much to discover about the nature and relative efficacy of specific techniques (Clark & Williamon, 2012) and how these might most effectively be incorporated into musical training (Wollner & Williamon, 2007). The current article describes and discusses an example of an imagery-based musical teaching method that specifically aims to facilitate memorization and execution of precomposed, notated music, which performers are required to translate into sound from memory during performance (Bailes, 2009).Musical imagery is a form of musical thought (Bailes, 2009), and is understood to be multimodal (Keller, 2012), including auditory, motor, and visual components. Musical imagery can occur offline (in the absence of overt performance) and online during performance (Bishop, Bailes, & Dean, 2013; Keller, 2012). In fact, as several authors have pointed out, performance necessarily includes imagery processes (Bernardi et al., 2013; Connolly & Williamon, 2004), generated either deliberately or in automatic response to internal or external cues (Bailes, 2009; Keller, 2012). We here define imagery as the deliberate internal generation of imagery in the absence of selfgenerated sensory feedback in the missing modality or modalities; imagery may therefore occur with or without a score, instrument or auditory model, and in the presence or absence of overt movement. Recent neuroscientific research points to both overlaps and differences in cognitive processing between and physical that may help to explain some of the effects of imagery rehearsal. …

Room Acoustical Parameters as Predictors of Room Acoustical Impression: What Do We Know and What Would We Like to Know?

Room acoustical parameters are audio features, usually extracted from monaural or binaural measurements of room acoustical environments, and used to predict different aspects of the ‘room acoustical impression’. The paper takes a closer look at the nature of this perceptional construct and at different approaches to develop a psychological measuring instrument for the multidimensional perceptional profile of room acoustical environments. Even after several decades of research, there is no satisfactory solution available for this purpose. The reasons for this lie in methodological deficits with respect to test development and item analysis as well as in a much too small sample of stimuli used in previous studies. Prospects for progress are opened up by state-of-the-art technologies for room acoustical simulation and auralization, which may be used to provide a large and representative sample of room acoustical environments as well as an authentic presentation in experimental studies. The fundamental perceptional components delivered by this approach will, most likely, not be predictable by traditional room acoustical parameters, but require advanced measurement techniques based on spherical arrays of transducers for both source and receiver characteristics, as well as new auditory models for feature extraction. The physical and psychological aspects of the problem are, in any case, inextricably linked with each other.

Tracking cortical entrainment in neural activity: auditory processes in human temporal cortex.

A primary objective for cognitive neuroscience is to identify how features of the sensory environment are encoded in neural activity. Current auditory models of loudness perception can be used to make detailed predictions about the neural activity of the cortex as an individual listens to speech. We used two such models (loudness-sones and loudness-phons), varying in their psychophysiological realism, to predict the instantaneous loudness contours produced by 480 isolated words. These two sets of 480 contours were used to search for electrophysiological evidence of loudness processing in whole-brain recordings of electro- and magneto-encephalographic (EMEG) activity, recorded while subjects listened to the words. The technique identified a bilateral sequence of loudness processes, predicted by the more realistic loudness-sones model, that begin in auditory cortex at ~80 ms and subsequently reappear, tracking progressively down the superior temporal sulcus (STS) at lags from 230 to 330 ms. The technique was then extended to search for regions sensitive to the fundamental frequency (F0) of the voiced parts of the speech. It identified a bilateral F0 process in auditory cortex at a lag of ~90 ms, which was not followed by activity in STS. The results suggest that loudness information is being used to guide the analysis of the speech stream as it proceeds beyond auditory cortex down STS toward the temporal pole.

Contralateral efferent suppression of human hearing sensitivity.

The present study aimed at characterizing the suppressing effect of contralateral medial olivocochlear (MOC) efferents on human auditory sensitivity and mechanical cochlear responses at sound levels near behavioral thresholds. Absolute thresholds for pure tones of 500 and 4000 Hz with durations between 10–500 ms were measured in the presence and in the absence of a contralateral broadband noise. The intensity of the noise was fixed at 60 dB SPL to evoke the contralateral MOC reflex without evoking the middle-ear muscle reflex. In agreement with previously reported findings, thresholds measured without the contralateral noise decreased with increasing tone duration, and the rate of decrease was faster at 500 than at 4000 Hz. Contralateral stimulation increased thresholds by 1.07 and 1.72 dB at 500 and 4000 Hz, respectively. The mean increase (1.4 dB) just missed statistical significance (p = 0.08). Importantly, the across-frequency mean threshold increase was significantly greater for long than for short probes. This effect was more obvious at 4000 Hz than at 500 Hz. Assuming that thresholds depend on the MOC-dependent cochlear mechanical response followed by an MOC-independent, post-mechanical detection mechanism, the present results at 4000 Hz suggest that MOC efferent activation suppresses cochlear mechanical responses more at lower than at higher intensities across the range of intensities near threshold, while the results at 500 Hz suggest comparable mechanical suppression across the threshold intensity range. The results are discussed in the context of central masking and of auditory models of efferent suppression of cochlear mechanical responses.

Fatigue Modeling via Mammalian Auditory System for Prediction of Noise Induced Hearing Loss.

Noise induced hearing loss (NIHL) remains as a severe health problem worldwide. Existing noise metrics and modeling for evaluation of NIHL are limited on prediction of gradually developing NIHL (GDHL) caused by high-level occupational noise. In this study, we proposed two auditory fatigue based models, including equal velocity level (EVL) and complex velocity level (CVL), which combine the high-cycle fatigue theory with the mammalian auditory model, to predict GDHL. The mammalian auditory model is introduced by combining the transfer function of the external-middle ear and the triple-path nonlinear (TRNL) filter to obtain velocities of basilar membrane (BM) in cochlea. The high-cycle fatigue theory is based on the assumption that GDHL can be considered as a process of long-cycle mechanical fatigue failure of organ of Corti. Furthermore, a series of chinchilla experimental data are used to validate the effectiveness of the proposed fatigue models. The regression analysis results show that both proposed fatigue models have high corrections with four hearing loss indices. It indicates that the proposed models can accurately predict hearing loss in chinchilla. Results suggest that the CVL model is more accurate compared to the EVL model on prediction of the auditory risk of exposure to hazardous occupational noise.

An Objective Analysis Method for Perceptual Quality of a Virtual Bass System

Due to the physical size and frequency response constraints of miniaturized and flat panel loudspeakers, low frequency reproduction from these loudspeakers is generally limited and unsatisfactory. The virtual bass system (VBS) enhances the bass performance of such loudspeakers by tricking the human brain to perceive the fundamental frequency from its higher harmonics. Problematically, the additional harmonics from VBS can also introduce perceptual distortion and reduce the audio quality. Therefore, a reliable method to assess the quality of VBS-enhanced signals is necessary in the designing of VBS. Since subjective experiments are often time-consuming and may be inconsistent, it is desirable to develop an objective assessment method for VBS. Earlier studies only utilized some simple objective metrics, which generally do not consider the human auditory model and are unable to accurately predict the perceptual quality of VBS. In this paper, we introduce a perceptual quality-assessment method for VBS based on the model output variables (MOVs) of the ITU Recommendation ITU-R BS.1387. Suitable combinations of MOVs are selected to derive perceptual quality metrics that correlate closely to the subjective quality. A verification experiment is presented to justify the accuracy of the metrics.

Speaker identification framework by peripheral and central auditory models

Speaker identification framework by peripheral and central auditory models

Predicting the perceived sound quality of frequency-compressed speech.

The performance of objective speech and audio quality measures for the prediction of the perceived quality of frequency-compressed speech in hearing aids is investigated in this paper. A number of existing quality measures have been applied to speech signals processed by a hearing aid, which compresses speech spectra along frequency in order to make information contained in higher frequencies audible for listeners with severe high-frequency hearing loss. Quality measures were compared with subjective ratings obtained from normal hearing and hearing impaired children and adults in an earlier study. High correlations were achieved with quality measures computed by quality models that are based on the auditory model of Dau et al., namely, the measure PSM, computed by the quality model PEMO-Q; the measure qc, computed by the quality model proposed by Hansen and Kollmeier; and the linear subcomponent of the HASQI. For the prediction of quality ratings by hearing impaired listeners, extensions of some models incorporating hearing loss were implemented and shown to achieve improved prediction accuracy. Results indicate that these objective quality measures can potentially serve as tools for assisting in initial setting of frequency compression parameters.

A bio-inspired feature extraction for robust speech recognition

In this paper, a feature extraction method for robust speech recognition in noisy environments is proposed. The proposed method is motivated by a biologically inspired auditory model which simulates the outer/middle ear filtering by a low-pass filter and the spectral behaviour of the cochlea by the Gammachirp auditory filterbank (GcFB). The speech recognition performance of our method is tested on speech signals corrupted by real-world noises. The evaluation results show that the proposed method gives better recognition rates compared to the classic techniques such as Perceptual Linear Prediction (PLP), Linear Predictive Coding (LPC), Linear Prediction Cepstral coefficients (LPCC) and Mel Frequency Cepstral Coefficients (MFCC). The used recognition system is based on the Hidden Markov Models with continuous Gaussian Mixture densities (HMM-GM).

Auditory-inspired pitch extraction using a synchrony capture filterbank

The question of how harmonic sounds in speech and music produce strong, low pitches at their fundamental frequencies, F0’s, has been of theoretical and practical interest to scientists and engineers for many decades. Currently the best auditory models for F0 pitch, (e.g., Meddis & Hewitt, 1991), are based on bandpass filtering (cochlear mechanics), half-wave rectification and low-pass filtering (hair cell transduction, synaptic transmission), channel autocorrelations (all-order interspike interval distributions) aggregated into a summary autocorrelation, followed by an analysis that determines the most prevalent interspike intervals. As a possible alternative to explicit autocorrelation computations, we propose an alternative model that uses an adaptive Synchrony Capture Filterbank (SCFB) in which channels in a filterbank neighborhood are driven exclusively (captured) by dominant frequency components closest to them. Channel outputs are then adaptively phase aligned with respect to a common time reference to compute a Summary Phase Aligned Function (SPAF), aggregated across all channels, from which F0 can then be easily extracted. Possible relations to brain rhythms and phase-locked loops are discussed. [Work supported by AFSOR FA9550-09-1-0119, Invited to special session about Synchronization in Musical Acoustics and Music Psychology.]

Computational modeling of individual differences in behavioral estimates of cochlear nonlinearities.

Temporal masking curves (TMCs) are often used to estimate cochlear compression in individuals with normal and impaired hearing. These estimates may yield a wide range of individual differences, even among subjects with similar quiet thresholds. This study used an auditory model to assess potential sources of variance in TMCs from 51 listeners in Poling et al. [J Assoc Res Otolaryngol, 13:91-108 (2012)]. These sources included threshold elevation, the contribution of outer and inner hair cell dysfunction to threshold elevation, compression of the off-frequency linear reference, and detection efficiency. Simulations suggest that detection efficiency is a primary factor contributing to individual differences in TMCs measured in normal-hearing subjects, while threshold elevation and the contribution of outer and inner hair cell dysfunction are primary factors in hearing-impaired subjects. Approximating the most compressive growth rate of the cochlear response from TMCs was achieved only in subjects with the highest detection efficiency. Simulations included off-frequency nonlinearity in basilar membrane and inner hair cell processing; however, this nonlinearity did not improve predictions, suggesting that other sources, such as the decay of masking and the strength of the medial olivocochlear reflex, may mimic off-frequency nonlinearity. Findings from this study suggest that sources of individual differences can play a strong role in behavioral estimates of compression, and these sources should be considered when using forward masking to study cochlear function in individual listeners or across groups of listeners.

Duifhuis pitch: neuromagnetic representation and auditory modeling.

When a high harmonic is removed from a cosine-phase harmonic complex, we hear a sine tone pop out of the perception; the sine tone has the pitch of the high harmonic, while the tone complex has the pitch of its fundamental frequency, f0. This phenomenon is commonly referred to as Duifhuis Pitch (DP). This paper describes, for the first time, the cortical representation of DP observed with magnetoencephalography. In experiment 1, conditions that produce the perception of a DP were observed to elicit a classic onset response in auditory cortex (P1m, N1m, P2m), and an increment in the sustained field (SF) established in response to the tone complex. Experiment 2 examined the effect of the phase spectrum of the complex tone on the DP activity: Schroeder-phase negative waves elicited a transient DP complex with a similar shape to that observed with cosine-phase waves but with much longer latencies. Following the transient DP activity, the responses of the negative and positive Schroeder-phase waves converged, and the increment in the SF slowly died away. In the absence of DP, the two Schroeder-phase conditions with low peak factors both produced larger SFs than cosine-phase waves with large peak factors. A model of the auditory periphery that includes coupling between adjacent frequency channels is used to explain the early neuromagnetic activity observed in auditory cortex.

Auditory Model Identification Using REVCOR Method

Auditory models are very useful in many applications such as speech coding and compression, cochlea prosthesis, and audio watermarking. In this paper we will develop a new auditory model based on the REVCOR method. This technique is based on the estimation of the impulse response of a suitable filter characterizing the auditory neuron and the cochlea. The first step of our study is focused on the development of a mathematical model based on the gammachirp system. This model is then programmed, implemented and simulated under Matlab. The obtained results are compared with the experimental values (REVCOR experiments) for the validation and a better optimization of the model parameters. Two objective criteria are used in order to optimize the audio model estimation which are the SNR (signal to noise ratio) and the MQE (mean quadratic error). The simulation results demonstrated that for the auditory model, only a reduced number of channels are excited (from 3 to 6). This result is very interesting for auditory implants because only significant channels will be stimulated. Besides, this simplifies the electronic implementation and medical intervention.