Equivalent Rectangular Bandwidth Research Articles

Automatic speaker and age identification of children from raw speech using sincNet over ERB scale

This paper presents the newly developed non-native children’s English speech (NNCES) corpus to reveal the findings of automatic speaker and age recognition from raw speech. Convolutional neural networks (CNN), which have the ability to learn low-level speech representations, can be fed directly with raw speech signals instead of using traditional hand-crafted features. Moreover, the filters that were learned using standard CNNs appeared to be noisy because they consider all elements of each filter. In contrast, sincNet can be able to generate more meaningful filters simply by replacing the first convolutional layer by a sinc-layer in standard CNNs. The low and high cutoff frequencies of the rectangular band-pass filter are the only parameters that can be learned in sincNet, which has the potential to extract significant speech cues from the speaker, such as pitch and formants. In this work, the sincNet model is significantly changed by switching from baseline Mel scale initializations to equivalent rectangular bandwidth (ERB) initializations, which has the added benefit of allocating additional filters in the lower region of the spectrum. Additionally, it needs to be highlighted that the novel sincNet model is well suited to identify the age of the children. The investigations on both read and spontaneous speech tasks in speaker identification, gender independent & dependent age-group identification of children outperform the baseline models with varying relative improvements in terms of accuracy.

Investigating the Performance of Gammatone Filters and Their Applicability to Design Cochlear Implant Processing System

Commercially available cochlear implants are designed to aid profoundly deaf people in understanding speech and environmental sounds. A typical cochlear implant uses a bank of bandpass filters to decompose an audio signal into a set of dynamic signals. These filters’ critical center frequencies f0 imitate the human cochlea’s vibration patterns caused by audio signals. Gammatone filters (GTFs), with two unique characteristics: (a) an appropriate “pseudo resonant” frequency transfer function, mimicking the human cochlea, and (b) realizing efficient hardware implementation, could demonstrate them as unique candidates for cochlear implant design. Although GTFs have recently attracted considerable attention from researchers, a comprehensive exposition of GTFs is still absent in the literature. This paper starts by enumerating the impulse response of GTFs. Then, the magnitude spectrum, |H(f)|, and bandwidth, more specifically, the equivalent rectangular bandwidth (ERB) of GTFs, are derived. The simulation results suggested that optimally chosen filter parameters, e.g., critical center frequencies,f0; temporal decay parameter, b; and order of the filter, n, can minimize the interference of the filter bank frequencies and very likely model the filter bandwidth (ERB), independent of f0b. Finally, these optimized filters are applied to delineate a filter bank for a cochlear implant design based on the Clarion processor model.

Array Configuration-Agnostic Personal Voice Activity Detection Based on Spatial Coherence

Personal voice activity detection (PVAD) has received increased attention due to the growing popularity of personal mobile devices and smart speakers. PVAD is often an integral element to speech enhancement and recognition for these applications in which lightweight signal processing is only enabled for the target user. However, in real-world scenarios, the detection performance may degrade because of competing speakers, background noise, and reverberation. To address this problem, we proposed to use equivalent rectangular bandwidth (ERB)-scaled spatial coherence as the input feature to train an array configuration-agnostic PVAD (ARCA-PVAD) network. Whereas the network model requires only 112k parameters, it exhibits excellent detection performance and robustness in adverse acoustic conditions. Notably, the proposed ARCA-PVAD system is scalable to array configurations (geometry, number of microphones, and spacing). Experimental results have demonstrated the superior performance achieved by the proposed ARCA-PVAD system over a baseline in terms of the area under receiver operating characteristic curve (AUC) and equal error rate (EER).

Distinguishing between straight and curved sounds: Auditory shape in pitch, loudness, and tempo gestures.

Sound-based trajectories or sound gestures draw links to spatiokinetic processes. For instance, a gliding, decreasing pitch conveys an analogous downward motion or fall. Whereas the gesture's pitch orientation and range convey its meaning and magnitude, respectively, the way in which pitch changes over time can be conceived of as gesture shape, which to date has rarely been studied in isolation. This article reports on an experiment that studied the perception of shape in uni-directional pitch, loudness, and tempo gestures, each assessed for four physical scalings. Gestures could increase or decrease over time and comprised different frequency and sound level ranges, durations, and different scaling contexts. Using a crossmodal-matching task, participants could reliably distinguish between pitch and loudness gestures and relate them to analogous visual line segments. Scalings based on equivalent-rectangular bandwidth (ERB) rate for pitch and raw signal amplitude for loudness were matched closest to a straight line, whereas other scalings led to perceptions of exponential or logarithmic curvatures. The investigated tempo gestures, by contrast, did not yield reliable differences. The reliable, robust perception of gesture shape for pitch and loudness has implications on various sound-design applications, especially those cases that rely on crossmodal mappings, e.g., visual analysis or control interfaces like audio waveforms or spectrograms. Given its perceptual relevance, auditory shape appears to be an integral part of sound gestures, while illustrating how crossmodal correspondences can underpin auditory perception.

Speech directivity patterns generated from a high-fidelity speech corpus

Human talkers are directional sound sources—a phenomenon that has consequences for speech perception in multi-talker environments. Directivity patterns for speech showing frequency- and angle-dependent radiation reveal that speech generally becomes more directional toward the front of the talker as frequency increases. Differences in physical attributes can lead to individual variability in directivity patterns across talkers. Here, we examine individual variability in speech directivity using frequency-dependentdirectivity indices and directivity maps. Speech directivity was examined in the horizontal plane using a corpus of simultaneous multi-channelfull-bandwidth (48-kHz sampling rate) recordings of the Bamford-Kowal-Bench (BKB) sentences recorded in an anechoic chamber. Thirty subjects (15 female) were recorded. The long-term average speech spectrum was utilized to calculate directivity indices in 1-ERB (equivalent rectangular bandwidth) bands. Gender differences in directivity indices were evaluated using a linear mixed-effects model. There was no main effect of gender. There was a main effect of ERB band with higher-frequency bands tending to have higher (i.e., more directional) directivity indices, however there was a nonmonotonic relationship between average directivity indices and frequency. Directivity maps demonstrated individual differences in speech radiation. [Work supported by NIH under Grant No. R01-DC019745.]

Supra-threshold deficits in normal hearing military recruits exposed to impulse noise.

The aim of this study was to determine the effect of impulse noise exposure on various proxy measures of cochlear synaptopathy in young military recruits. A total of 27 military recruits with exposure to firearm and artillery noise and 13 non exposed participants were recruited. All presented with normal hearing thresholds and the presence of distortion product otoacoustic emissions (DPOAEs). The Noise Exposure Structured Interview (NESI) was used to quantify noise exposure. Speech perception in noise (SPiN), equivalent rectangular bandwidth (ERB) of auditory filters, auditory brainstem response wave I amplitude, wave I amplitude growth function, wave I/V amplitude ratio, wave V latency, wave V latency shift with ipsilateral noise, and the summating potential/action potential ratio of the electrocochleography were measured. In military participants, SPiN was worse, ERB at 4 kHz was larger, wave I amplitude at 75 dBnHL was reduced, and wave V latency was delayed. However, no significant correlations were observed between NESI and auditory measures, once multiplicity of tests was controlled for. These results suggest that military recruits may exhibit supra-threshold deficits, despite presenting with normal hearing thresholds and presence of DPOAEs. Future studies should include a measure of auditory filters in their test battery.

Multi-Scale Recurrence Quantification Measurements for Voice Disorder Detection

Due to the complexity and non-stationarity of the voice generation system, the nonlinearity of speech signals cannot be accurately quantified. Recently, the recurrence quantification analysis method has been used for voice disorder detection. In this paper, multiscale recurrence quantification measures (MRQMs) are proposed. The signals are reconstructed in the high-dimensional phase space at the equivalent rectangular bandwidth scale. Recurrence plots (RPs) combining the characteristics of human auditory perception are drawn with an appropriate recurrence threshold. Based on the above, the nonlinear dynamic recurrence features of the speech signal are quantized from the recurrence plot of each frequency channel. Furthermore, this paper explores the recurrence quantification thresholds that are most suitable for pathological voices. Our results show that the proposed MRQMs with support vector machine (SVM), random forest (RF), Bayesian network (BN) and Local Weighted Learning (LWL) achieve an average accuracy of 99.45%, outperforming traditional features and other complex measurements. In addition, MRQMs also have the potential for multi-classification of voice disorder, achieving an accuracy of 89.05%. This study demonstrates that MRQMs can characterize the recurrence characteristic of pathological voices and effectively detect voice disorders.

Auditory filter shapes derived from forward and simultaneous masking at low frequencies: Implications for human cochlear tuning

Behavioral forward-masking thresholds with a spectrally notched-noise masker and a fixed low-level probe tone have been shown to provide accurate estimates of cochlear tuning. Estimates using simultaneous masking are similar but generally broader, presumably due to nonlinear cochlear suppression effects. So far, estimates with forward masking have been limited to frequencies of 1 kHz and above. This study used spectrally notched noise under forward and simultaneous masking to estimate frequency selectivity between 200 and 1000 Hz for young adult listeners with normal hearing. Estimates of filter tuning at 1000 Hz were in agreement with previous studies. Estimated tuning broadened below 1000 Hz, with the filter quality factor based on the equivalent rectangular bandwidth (QERB) decreasing more rapidly with decreasing frequency than predicted by previous equations, in line with earlier predictions based on otoacoustic-emission latencies. Estimates from simultaneous masking remained broader than those from forward masking by approximately the same ratio. The new data provide a way to compare human cochlear tuning estimates with auditory-nerve tuning curves from other species across most of the auditory frequency range.

Diagnosis of Depression Based on Four-Stream Model of Bi-LSTM and CNN From Audio and Text Information

Recent development trends in artificial intelligence applications have seen increasing interest in the design of automated systems for depression detection and diagnosis among the affective computing community. Particularly, active research has been conducted in depression diagnosis, based on multi-modal approaches in deep learning technology, which enable utilization of various information through fusion of varied data types. This study proposes a four-stream-based depression diagnosis model consisting of Bidirectional Long Short-Term Memory (Bi-LSTM) and convolutional neural networks (CNN), using speech and text data. One-dimensional features of audio signals are extracted using Mel Frequency Cepstral Coefficients and Gammatone Cepstral Coefficients, and two-dimensional features are extracted from Bark, equivalent rectangular bandwidth, and Log-Mel spectrograms, based on time-frequency transform. The extracted features are applied to Bi-LSTM and CNN-based transfer learning models. Word encoding was used for mapping of text to sequences with numeric indices, and word embedding used for representation of all words in numeric dense vectors. These were applied to Bi-LSTM and n-gram-based CNN models. Finally, an ensemble of the softmax values output from the four deep learning models was used to perform depression diagnosis, based on the proposed four-stream model. Using the proposed model, experiments were performed with the Extended Distress Analysis Interview Corpus Wizard of Oz depression database and other datasets. Experimental results showed improved performance by 10.7% to 11.9% over two-stream-based state-of-the-art methods. This demonstrates that the proposed model is effective for depression diagnosis.

Voice to Face Recognition Using Spectral ERB-DMLP Algorithms

Designing an authentication system for securing the power plants are important to allow only specific staffs of the power plant to access the certain blocks so that they can be restricted from using high risk-oriented equipment. This authentication is also vital to prevent any security threats or risks like compromises of business server, release of confidential data etc. Though conventional works attempted to accomplish better authentication, they lacked with respect to accuracy. Hence, the study aims to enhance the recognition rate by introducing a voice recognition system as a personal authentication based on Deep Learning (DL) due to its ability to perform effective learning. The study proposes Equivalent Rectangular Bandwidth and Deep Multi-Layer Perceptron (ERB-DMLP) as it has the ability to perform efficient and relevant feature extraction and faster classification. This algorithm also has the ability to establish effective correlation between voices and images and achieve the semantic relationship between them. Voice pre-processing is initially performed to make it suitable for further processing by removing the noise and enhancing the quality of signal. This process is also vital to minimize the extra computations so that the overall efficacy of the system can be made flexible by considering the audio files as features and the images as labels to identify a person’s voice by classifying the extracted features from the ERB Feature Extraction. This is then passed as the input into DMLP model to classify the persons, and trained the model to make an accurate classification of audio with corresponding image labels, and perform the performance test based on the trained model. Flexibility, relevant feature extraction and faster classification ability of the proposed work has made it explore better outcomes that is confirmed through results.

Using ideal binary masking based on signal-to-noise ratio of temporal amplitude envelope to improve the intelligibility of speech in noise

Ideal Binary Masking (IBM), using prior information, improves speech intelligibility by attenuating noisy components with a scaling factor applied to the noise. The main challenge is to construct an appropriate decision-making model to identify noise- or speech- dominant components. In this study, we utilized the signal-to-noise ratio (SNR) of the temporal amplitude envelope in the frequency-time domain. We firstly divided the noisy speech from 200 Hz to 6 kHz, processed by MATLAB, into 16 contiguous subbands each with bandwidth approximately 1.5 times an equivalent rectangular bandwidth. The subband envelopes were produced by means of the absolute value of the signal. SNRs of the temporal envelope were calculated for 40 ms windows. The mask was unity when the SNR was greater than −5dB; otherwise, it was 0.5. We evaluated the performance of the proposed IBM on word scores obtained with different speech in speech-spectrum shaped noise SNR values of −2, −4, −6, and −8 dB. Sixteen native speakers (age 28 ± 3 years) with normal hearing were recruited for the study and underwent the Modified Rhyme Test to assess intelligibility. Statistically significant increases of up to 20% in mean word scores were obtained by this IBM. [Work supported by NIOSH.]

Automatic responses to musical intervals: Contrasts in acoustic roughness predict affective priming in Western listeners.

The aim of the present study is to determine which acoustic components of harmonic consonance and dissonance influence automatic responses in a simple cognitive task. In a series of affective priming experiments, eight pairs of musical intervals were used to measure the influence of acoustic roughness and harmonicity on response times in a word-classification task conducted online. Interval pairs that contrasted in roughness induced a greater degree of affective priming than pairs that did not contrast in terms of their roughness. Contrasts in harmonicity did not induce affective priming. A follow-up experiment used detuned intervals to create higher levels of roughness contrasts. However, the detuning did not lead to any further increase in the size of the priming effect. More detailed analysis suggests that the presence of priming in intervals is binary: in the negative primes that create congruency effects the intervals' fundamentals and overtones coincide within the same equivalent rectangular bandwidth (i.e., the minor and major seconds). Intervals that fall outside this equivalent rectangular bandwidth do not elicit priming effects, regardless of their dissonance or negative affect. The results are discussed in the context of recent developments in consonance/dissonance research and vocal similarity.

Noise thresholds in harmonic series maskers.

The presence of noise is a salient cue to the perception of breathiness and aspiration in speech sounds. The detection of noise within harmonic series (maskers) composed of unresolved components was found to depend on the fundamental frequency (fo) and the overall level of the masker [Gockel, Moore, and Patterson (2002). J. Acoust. Soc. Am., 111 (6), 2759-2770]. In the present study, noise detection thresholds were measured as a function of the frequency range, the fo, and the overall level of harmonic maskers. Frequency range was specified in equivalent rectangular bandwidth (ERB) units (3-13, 13-23, 23-33, or 3-33 ERBs). The results were consistent with the idea that listeners rely on spectral cues when maskers comprise only resolved components (3-13 ERBs), and on temporal (dip listening) cues when maskers contain only unresolved components (23-33 ERBs). Noise detection thresholds were generally lower when masker level was high (70 dBA) than when it was low (50 dBA). Masker fo affected thresholds only when listeners relied on spectral cues for noise detection. With the wideband (3-33 ERBs) masker, listeners likely detected noise by focusing on the frequency band (23-33 ERBs) with the most advantageous noise-to-harmonic ratio.

Estimating a Sleep Apnea Hypopnea Index Based on the ERB Correlation Dimension of Snore Sounds.

This paper proposes a new perspective of analyzing non-linear acoustic characteristics of the snore sounds. According to the ERB (Equivalent Rectangular Bandwidth) scale used in psychoacoustics, the ERB correlation dimension (ECD) of the snore sound was computed to feature different severity levels of sleep apnea hypopnea syndrome (SAHS). For the training group of 93 subjects, snore episodes were manually segmented and the ECD parameters of the snores were extracted, which established the gaussian mixture models (GMM). The nocturnal snore sound of the testing group of another 120 subjects was tested to detect SAHS snores, thus estimating the apnea hypopnea index (AHI), which is called AHIECD. Compared to the AHIPSG value of the gold standard polysomnography (PSG) diagnosis, the estimated AHIECD achieved an accuracy of 87.5% in diagnosis the SAHS severity levels. The results suggest that the ECD vectors can be effective parameters for screening SAHS.

Recognition of Pathological Voices by Human Factor Cepstral Coefficients (HFCC)

Human speech is a means of communication that is very important in our daily lives. It is characterized by its great ability to transmit our ideas, our emotions, our personality etc. So, any alteration of the voice can prevent the person from exercising his professional and daily life naturally. It is for these reasons that it is very necessary to implement systems for detecting and classifying vocal pathologies. These automatic systems can help clinicians customize and detect the existence of any vocal pathology. In this context, several tools have been introduced to achieve early detection of voice disorders. Among these tools are the Human Factor Cepstral Coefficients (HFCC) combined with prosodic parameters, the Noise-Harmonic Ratio (NHR), the Harmonic-Noise Ratio (HNR), analysis of trend Fluctuations (DFA) and Fundamental frequency (F0). These parameters are introduced and calculated in every frame. In this study, we used a variation of HFCC called Equivalent Rectangular Bandwidth (ERB) to study the effects of HFCC on the classification of pathological voices. Using the HTK classifiers, the classification is carried out on two pathological databases, Massachusetts Eye and Ear Infirmary (MEEI) and Saarbruecken Voice Database (SVD). To assess the performance of the system, we used sensitivity and specificity.

Speech emotion recognition using cepstral features extracted with novel triangular filter banks based on bark and ERB frequency scales

Speech emotion recognition (SER) refers to the process of recognizing the emotional state of the speaker from the speech utterance. In earlier studies, wide varieties of cepstral features have been proposed to develop SER systems. The mel-frequency cepstral coefficients (MFCC), and human-factor cepstral coefficients (HFCC) are two popularly used variants of cepstral features. MFCC and HFCC features are extracted from speech signals using mel and human-factor filter banks, respectively. The magnitude response of individual filters in these filter banks is triangular in shape. As a result, these filter banks are referred to as triangular filter banks (TFB) and the corresponding extracted cepstral coefficients can be denoted as TFBCC-M (in case of MFCC) and TFBCC-HF (in case of HFCC). The mel-filter bank (TFB-M) is constructed using mel-scale, while the human-factor filter bank (TFB-HF) is constructed using human factor scale, which is a combination of mel and equivalent rectangular bandwidth (ERB) scales. Similarly, different frequency scales can be used to realize different TFBs to extract different types of TFBCC features. In this direction, this paper proposes two new TFBs denoted as TFB-B and TFB-E, realized using bark and ERB scales to extract new cepstral features referred to as TFBCC-B and TFBCC-E, respectively. The mathematical background to construct the proposed TFB-B and TFB-E is presented. The proposed filter banks are used along with the conventional TFB-M and TFB-HF to extract four different types of TFBCC features. These features are extracted from the emotional speech signals of two databases, namely Berlin database of emotional speech (Emo-DB) and Surrey audio-visual expressed emotion speech database (SAVEE). The extracted features are used to develop speaker-dependent (SD) and speaker-independent (SI) based SER systems using support vector machines. The performance of the respective features is analyzed in terms of isolated and combined usage. The experimental results show that the cepstral features extracted using the proposed TFBs are effective in characterizing and recognizing emotions similar to conventional MFCC and HFCC features. Moreover, the combined use of different cepstral features have resulted to improve the overall recognition performance of SER systems. In case of Emo-DB database, isolated use of the proposed TFBCC-B and TFBCC-E features achieve recognition accuracies of 83.23% and 81.99% for SD scenario, and 75% and 60.94% for SI scenario, respectively. Similarly, for SAVEE database, the recognition accuracies of 75% and 66.67% for SD scenario, and 44.17% and 55% for SI scenario are achieved. In case of Emo-DB database, the maximum recognition accuracies of 86.96% (for different combinations of conventional and proposed features namely, TFBCC-{(M+E), (M+B+E), (HF+B+E), (M+HF+B+E)}) and 77.08% (for combination TFBCC-(M+B+E)) are achieved for SD and SI scenarios, respectively. Similarly, for SAVEE database, the maximum recognition accuracies of 77.08% (for combination TFBCC-(M+HF+E)), and 55.83% (for combination TFBCC-(B+E)) are achieved for SD and SI scenarios, respectively.

Twin identification from speech: linear and non-linear cepstral features and models

Identifying a speaker from the speech by using computer-based learning techniques is normally in vogue for many years in research perspective. Twin identification from visual features is a challenging work specifically to identify a twin from identical twin pairs because they look alike in exhibiting manners, habits and so on. Speech can be used as a biometric to perform twin identification by imbibing adequate training imparted considering the fact that their vocal tract structure would be different in producing speech sounds. This work mainly highlights the usage of linear and non-linear frequency based Cepstral features and modelling methods to estimate the performance of the twin identification system. Set of speech utterances are separated into two sets namely training and testing. Utterances considered for training are concatenated and by performing the conventional pre-processing and appropriate feature extraction techniques, the proposed features are extracted. The feature vectors are normalized and given to the modeling techniques to create templates. Proposed features extracted from the utterances considered for testing are applied to the models and based on the classification criteria, twin is identifiedand the accuracy is categorized as sub-optimal and true success rates. Perceptual features with filters spaced in Equivalent Rectangular Bandwidth (ERB) for critical band provides better accuracy for the individual classification and decision level fusion classification.

Comparison of distortion-product otoacoustic emission and stimulus-frequency otoacoustic emission two-tone suppression in humans

Distortion-product otoacoustic emission (DPOAE) and stimulus-frequency otoacoustic emission (SFOAE) are two types of acoustic signals emitted by the inner ear in response to tonal stimuli. The levels of both emission types may be reduced by the inclusion of additional (suppressor) tones with the stimulus. Comparison of two-tone suppression properties across emission type addresses a clinically relevant question of whether these two types of emission provide similar information about cochlear status. The purpose of this study was to compare DPOAE suppression to SFOAE suppression from the same ear in a group of participants with normal hearing. Probe frequency was approximately 1000 Hz, and the suppressor frequency varied from -1.5 to 0.5 octaves relative to the probe frequency. DPOAE and SFOAE suppression were compared in terms of (1) suppression growth rate (SGR), (2) superimposed suppression tuning curves (STCs), and (3) STC-derived metrics, such as high-frequency slope, cochlear amplifier gain, and QERB (ERB, equivalent rectangular bandwidth). Below the probe frequency, the SGR was slightly greater than one for SFOAEs and slightly less than two for DPOAEs. There were no differences in STC metrics across emission types. These observations may provide useful constraints on physiology-based models of otoacoustic emission suppression.

Ripple spectrum resolution by prelingual and postlingual cochlear implant users

The correlation between ripple spectrum resolution and speech recognition was investigated for prelingual and postlingual cochlear implant users. The test signal had a band limited rippled spectrum (spectrum with alternated maxima and minima of spectral power) with equivalent rectangular bandwidth of 1 oct. The central frequency of the spectrum was 1, 2, or 4 kHz. The principle of rippled-spectrum test was to find the maximum ripple density (ripples/octave) at which listeners could detect the phase reversion of the spectrum. A balanced test set of Russian worlds by Greenberg-Zinder was used for estimation of speech recognition. The average ripple discrimination thresholds were approximately 2 ripples/octave in both pre- and postlingual patients. The thresholds did not depend from central frequency. The best correlation between speech recognition and rippled spectrum resolution was found in prelingual patients with perimodiolar electrode array compared with straight array (r = 0.85). In postlingual patients, correlation was markedly lower (r = 0.44 and r = 0.13 for rippled-spectrum tests centered at 1 or 2 kHz, respectively). [Work supported by the Russian Science Foundation, Grant No. 16-15-10046.]The correlation between ripple spectrum resolution and speech recognition was investigated for prelingual and postlingual cochlear implant users. The test signal had a band limited rippled spectrum (spectrum with alternated maxima and minima of spectral power) with equivalent rectangular bandwidth of 1 oct. The central frequency of the spectrum was 1, 2, or 4 kHz. The principle of rippled-spectrum test was to find the maximum ripple density (ripples/octave) at which listeners could detect the phase reversion of the spectrum. A balanced test set of Russian worlds by Greenberg-Zinder was used for estimation of speech recognition. The average ripple discrimination thresholds were approximately 2 ripples/octave in both pre- and postlingual patients. The thresholds did not depend from central frequency. The best correlation between speech recognition and rippled spectrum resolution was found in prelingual patients with perimodiolar electrode array compared with straight array (r = 0.85). In postlingual patient...

Acoustic coordinated reset therapy for tinnitus with perceptually relevant frequency spacing and levels

Acoustic coordinated reset (CR) therapy based on neuromodulation and neuroplasticity principles has been proposed for the treatment of tonal tinnitus. The original therapy involved periodic delivery of randomly ordered sequences of four low-level tones centered around the frequency of a tone that matched the tinnitus pitch, fT, with fixed ratios relative to fT and delivered several hours/day over several weeks. Here we transform the original CR tone selection method to a more perceptually-relevant equivalent rectangular bandwidth (ERB) frequency scale, the ERBN-number scale. Specifically, we provide a mathematical model that enables calculation of CR tones that accounts for fT- and hearing loss-related ERB widening and ERB overlaps and gaps of CR tone alignments. Further, the model ensures symmetric CR tone alignments based on modelling studies that indicate the effect is optimal if the CR stimuli are symmetrically spaced relative to the tinnitus-related population of abnormally synchronized cortical neurons to activate the adjacent sub-populations. We also present experimentally testable ERB-based CR tone alignment strategies and explain how to use the ERB-based model in experiments, clinical studies, other types of tinnitus sound treatment such as tailor-made notch music training and limitations of our approach.