Improved tactile speech robustness to background noise with a dual-path recurrent neural network noise-reduction method.

Many people with hearing loss struggle to understand speech in noisy environments, making noise robustness critical for hearing-assistive devices. Recently developed haptic hearing aids, which convert audio to vibration, can improve speech-in-noise performance for cochlear implant (CI) users and assist those unable to access hearing-assistive devices. They are typically body-worn rather than head-mounted, allowing additional space for batteries and microprocessors, and so can deploy more sophisticated noise-reduction techniques. The current study assessed whether a real-time-feasible dual-path recurrent neural network (DPRNN) can improve tactile speech-in-noise performance. Audio was converted to vibration on the wrist using a vocoder method, either with or without noise reduction. Performance was tested for speech in a multi-talker noise (recorded at a party) with a 2.5-dB signal-to-noise ratio. An objective assessment showed the DPRNN improved the scale-invariant signal-to-distortion ratio by 8.6 dB and substantially outperformed traditional noise-reduction (log-MMSE). A behavioural assessment in 16 participants showed the DPRNN improved tactile-only sentence identification in noise by 8.2%. This suggests that advanced techniques like the DPRNN could substantially improve outcomes with haptic hearing aids. Low-cost haptic devices could soon be an important supplement to hearing-assistive devices such as CIs or offer an alternative for people who cannot access CI technology.

An online implementation of a measure of spectro-temporal processing by cochlear-implant listeners.

The spectro-temporal ripple for investigating processor effectiveness (STRIPES) test is a psychophysical measure of spectro-temporal resolution in cochlear-implant (CI) listeners. It has been validated using direct-line input and loudspeaker presentation with listeners of the Advanced Bionics CI. This article investigates the suitability of an online application using wireless streaming (webSTRIPES) as a remote test. It reports a strong across-listener correlation between STRIPES thresholds obtained using laboratory testing with loudspeaker presentation vs remote testing with streaming presentation, with no significant difference in STRIPES thresholds between the two measures. WebSTRIPES also produced comparable and robust thresholds with users of the Cochlear CI.

SpeedCAP: An Efficient Method for Estimating Neural Activation Patterns Using Electrically Evoked Compound Action-Potentials in Cochlear Implant Users.

OBJECTIVES: Electrically evoked compound action-potentials (ECAPs) can be recorded using the electrodes in a cochlear implant (CI) and represent the synchronous responses of the electrically stimulated auditory nerve. ECAPs can be obtained using a forward-masking method that measures the neural response to a probe and masker electrode separately and in combination. The panoramic ECAP (PECAP) analyses measured ECAPs obtained using multiple combinations of masker and probe electrodes and uses a nonlinear optimization algorithm to estimate current spread from each electrode and neural health along the cochlea. However, the measurement of ECAPs from multiple combinations of electrodes is too time consuming for use in clinics. Here, we propose and evaluate SpeedCAP, a speedy method for obtaining the PECAP measurements that minimizes recording time by exploiting redundancies between multiple ECAP measures. DESIGN: In the first study, 11 users of Cochlear Ltd. CIs took part. ECAPs were recorded using the forward-masking artifact-cancelation technique at the most comfortable loudness level (MCL) for every combination of masker and probe electrodes for all active electrodes in the users' MAPs, as per the standard PECAP recording paradigm. The same current levels and recording parameters were then used to collect ECAPs in the same users with the SpeedCAP method. The ECAP amplitudes were then compared between the two conditions, as were the corresponding estimates of neural health and current spread calculated using the PECAP method previously described by Garcia et al. The second study measured SpeedCAP intraoperatively in 8 CI patients and with all maskers and probes presented at the same current level to assess feasibility. ECAPs for the subset of conditions where the masker and probe were presented on the same electrode were compared with those obtained using the slower approach leveraged by the standard clinical software. RESULTS: Data collection time was reduced from ≈45 to ≈8 minutes. There were no significant differences between normalized root mean squared error (RMSE) repeatability metrics for post-operative PECAP and SpeedCAP data, nor for the RMSEs calculated between PECAP and SpeedCAP data. The comparison achieved 80% power to detect effect sizes down to 8.2% RMSE. When between-participant differences were removed, both the neural-health (r = 0.73) and current-spread (r = 0.65) estimates were significantly correlated ( p < 0.0001, df = 218) between SpeedCAP and PECAP conditions across all electrodes, and showed RMSE errors of 12.7 ± 4.7% and 16.8 ± 8.8%, respectively (with the ± margins representing 95% confidence intervals). Valid ECAPs were obtained in all patients in the second study, demonstrating intraoperative feasibility of SpeedCAP. No significant differences in RMSEs were detectable between post- and intra-operative ECAP measurements, with the comparison achieving 80% power to detect effect sizes down to 13.3% RMSE. CONCLUSIONS: The improved efficiency of SpeedCAP provides time savings facilitating multi-electrode ECAP recordings in routine clinical practice. SpeedCAP data collection is sufficiently quick to record intraoperatively, and adds no more than 8.2% error to the ECAP amplitudes. Such measurements could thereafter be submitted to models such as PECAP to provide patient-specific patterns of neural activation to inform programming of clinical MAPs and identify causes of poor performance at the electrode-nerve interface of CI users. The speed and accuracy of these measurements also opens up a wide range of additional research questions to be addressed.

Helping People Hear Better with "Smart" Hearing Devices.

Millions of people around the world have difficulty hearing. Hearing aids and cochlear implants help people hear better, especially in quiet places. Unfortunately, these devices do not always help in noisy situations like busy classrooms or restaurants. This means that a person with hearing loss may struggle to follow a conversation with friends or family and may avoid going out. We used methods from the field of artificial intelligence to develop "smart" hearing aids and cochlear implants that can get rid of background noise. We play many different sounds into a computer program, which learns to pick out the speech sounds and filter out unwanted background noises. Once the computer program has been trained, it is then tested on new examples of noisy speech and can be incorporated into hearing aids or cochlear implants. These "smart" approaches can help people with hearing loss understand speech better in noisy situations.

From Microphone to Phoneme: An End-to-End Computational Neural Model for Predicting Speech Perception With Cochlear Implants.

GOAL: Advances in computational models of biological systems and artificial neural networks enable rapid virtual prototyping of neuroprostheses, accelerating innovation in the field. Here, we present an end-to-end computational model for predicting speech perception with cochlear implants (CI), the most widely-used neuroprosthesis. METHODS: The model integrates CI signal processing, a finite element model of the electrically-stimulated cochlea, and an auditory nerve model to predict neural responses to speech stimuli. An automatic speech recognition neural network is then used to extract phoneme-level speech perception from these neural response patterns. RESULTS: Compared to human CI listener data, the model predicts similar patterns of speech perception and misperception, captures between-phoneme differences in perceptibility, and replicates effects of stimulation parameters and noise on speech recognition. Information transmission analysis at different stages along the CI processing chain indicates that the bottleneck of information flow occurs at the electrode-neural interface, corroborating studies in CI listeners. CONCLUSION: An end-to-end model of CI speech perception replicated phoneme-level CI speech perception patterns, and was used to quantify information degradation through the CI processing chain. SIGNIFICANCE: This type of model shows great promise for developing and optimizing new and existing neuroprostheses.

Cochlear Implant Research and Development in the Twenty-first Century: A Critical Update.

Cochlear implants (CIs) are the world's most successful sensory prosthesis and have been the subject of intense research and development in recent decades. We critically review the progress in CI research, and its success in improving patient outcomes, from the turn of the century to the present day. The review focuses on the processing, stimulation, and audiological methods that have been used to try to improve speech perception by human CI listeners, and on fundamental new insights in the response of the auditory system to electrical stimulation. The introduction of directional microphones and of new noise reduction and pre-processing algorithms has produced robust and sometimes substantial improvements. Novel speech-processing algorithms, the use of current-focusing methods, and individualised (patient-by-patient) deactivation of subsets of electrodes have produced more modest improvements. We argue that incremental advances have and will continue to be made, that collectively these may substantially improve patient outcomes, but that the modest size of each individual advance will require greater attention to experimental design and power. We also briefly discuss the potential and limitations of promising technologies that are currently being developed in animal models, and suggest strategies for researchers to collectively maximise the potential of CIs to improve hearing in a wide range of listening situations.

The effect of increased channel interaction on speech perception with cochlear implants.

Cochlear implants (CIs) are neuroprostheses that partially restore hearing for people with severe-to-profound hearing loss. While CIs can provide good speech perception in quiet listening situations for many, they fail to do so in environments with interfering sounds for most listeners. Previous research suggests that this is due to detrimental interaction effects between CI electrode channels, limiting their function to convey frequency-specific information, but evidence is still scarce. In this study, an experimental manipulation called spectral blurring was used to increase channel interaction in CI listeners using Advanced Bionics devices with HiFocus 1J and MS electrode arrays to directly investigate its causal effect on speech perception. Instead of using a single electrode per channel as in standard CI processing, spectral blurring used up to 6 electrodes per channel simultaneously to increase the overlap between adjacent frequency channels as would occur in cases with severe channel interaction. Results demonstrated that this manipulation significantly degraded CI speech perception in quiet by 15% and speech reception thresholds in babble noise by 5 dB when all channels were blurred by a factor of 6. Importantly, when channel interaction was increased just on a subset of electrodes, speech scores were mostly unaffected and were only significantly degraded when the 5 most apical channels were blurred. These apical channels convey information up to 1 kHz at the apical end of the electrode array and are typically located at angular insertion depths of about 250 up to 500°. These results confirm and extend earlier findings indicating that CI speech perception may not benefit from deactivating individual channels along the array and that efforts should instead be directed towards reducing channel interaction per se and in particular for the most-apical electrodes. Hereby, causal methods such as spectral blurring could be used in future research to control channel interaction effects within listeners for evaluating compensation strategies.

The Panoramic ECAP Method: Estimating Patient-Specific Patterns of Current Spread and Neural Health in Cochlear Implant Users.

The knowledge of patient-specific neural excitation patterns from cochlear implants (CIs) can provide important information for optimizing efficacy and improving speech perception outcomes. The Panoramic ECAP ('PECAP') method (Cosentino et al. 2015) uses forward-masked electrically evoked compound action-potentials (ECAPs) to estimate neural activation patterns of CI stimulation. The algorithm requires ECAPs be measured for all combinations of probe and masker electrodes, exploiting the fact that ECAP amplitudes reflect the overlapping excitatory areas of both probes and maskers. Here we present an improved version of the PECAP algorithm that imposes biologically realistic constraints on the solution, that, unlike the previous version, produces detailed estimates of neural activation patterns by modelling current spread and neural health along the intracochlear electrode array and is capable of identifying multiple regions of poor neural health. The algorithm was evaluated for reliability and accuracy in three ways: (1) computer-simulated current-spread and neural-health scenarios, (2) comparisons to psychophysical correlates of neural health and electrode-modiolus distances in human CI users, and (3) detection of simulated neural 'dead' regions (using forward masking) in human CI users. The PECAP algorithm reliably estimated the computer-simulated scenarios. A moderate but significant negative correlation between focused thresholds and the algorithm's neural-health estimates was found, consistent with previous literature. It also correctly identified simulated 'dead' regions in all seven CI users evaluated. The revised PECAP algorithm provides an estimate of neural excitation patterns in CIs that could be used to inform and optimize CI stimulation strategies for individual patients in clinical settings.

An Instrumented Cochlea Model for the Evaluation of Cochlear Implant Electrical Stimulus Spread.

Cochlear implants use electrical stimulation of the auditory nerve to restore the sensation of hearing to deaf people. Unfortunately, the stimulation current spreads extensively within the cochlea, resulting in "blurring" of the signal, and hearing that is far from normal. Current spread can be indirectly measured using the implant electrodes for both stimulating and sensing, but this provides incomplete information near the stimulating electrode due to electrode-electrolyte interface effects. Here, we present a 3D-printed "unwrapped" physical cochlea model with integrated sensing wires. We integrate resistors into the walls of the model to simulate current spread through the cochlear bony wall, and "tune" these resistances by calibration with an in-vivo electrical measurement from a cochlear implant patient. We then use this model to compare electrical current spread under different stimulation modes including monopolar, bipolar and tripolar configurations. Importantly, a trade-off is observed between stimulation amplitude and current focusing among different stimulation modes. By combining different stimulation modes and changing intracochlear current sinking configurations in the model, we explore this trade-off between stimulation amplitude and focusing further. These results will inform clinical strategies for use in delivering speech signals to cochlear implant patients.

The Effect of Free-Field Presentation and Processing Strategy on a Measure of Spectro-Temporal Processing by Cochlear-Implant Listeners.

The STRIPES (Spectro-Temporal Ripple for Investigating Processor EffectivenesS) test is a psychophysical test of spectro-temporal resolution developed for cochlear-implant (CI) listeners. Previously, the test has been strictly controlled to minimize the introduction of extraneous, nonspectro-temporal cues. Here, the effect of relaxing many of those controls was investigated to ascertain the generalizability of the STRIPES test. Preemphasis compensation was removed from the STRIPES stimuli, the test was presented over a loudspeaker at a level similar to conversational speech and above the automatic gain control threshold of the CI processor, and listeners were tested using the everyday setting of their clinical devices. There was no significant difference in STRIPES thresholds measured across conditions for the 10 CI listeners tested. One listener obtained higher (better) thresholds when listening with their clinical processor. An analysis of longitudinal results showed excellent test-retest reliability of STRIPES over multiple listening sessions with similar conditions. Overall, the results show that the STRIPES test is robust to extraneous cues, and that thresholds are reliable over time. It is sufficiently robust for use with different processing strategies, free-field presentation, and in nonresearch settings.

Using Spectral Blurring to Assess Effects of Channel Interaction on Speech-in-Noise Perception with Cochlear Implants.

Cochlear implant (CI) listeners struggle to understand speech in background noise. Interactions between electrode channels due to current spread increase the masking of speech by noise and lead to difficulties with speech perception. Strategies that reduce channel interaction therefore have the potential to improve speech-in-noise perception by CI listeners, but previous results have been mixed. We investigated the effects of channel interaction on speech-in-noise perception and its association with spectro-temporal acuity in a listening study with 12 experienced CI users. Instead of attempting to reduce channel interaction, we introduced spectral blurring to simulate some of the effects of channel interaction by adjusting the overlap between electrode channels at the input level of the analysis filters or at the output by using several simultaneously stimulated electrodes per channel. We measured speech reception thresholds in noise as a function of the amount of blurring applied to either all 15 electrode channels or to 5 evenly spaced channels. Performance remained roughly constant as the amount of blurring applied to all channels increased up to some knee point, above which it deteriorated. This knee point differed across listeners in a way that correlated with performance on a non-speech spectro-temporal task, and is proposed here as an individual measure of channel interaction. Surprisingly, even extreme amounts of blurring applied to 5 channels did not affect performance. The effects on speech perception in noise were similar for blurring at the input and at the output of the CI. The results are in line with the assumption that experienced CI users can make use of a limited number of effective channels of information and tolerate some deviations from their everyday settings when identifying speech in the presence of a masker. Furthermore, these findings may explain the mixed results by strategies that optimized or deactivated a small number of electrodes evenly distributed along the array by showing that blurring or deactivating one-third of the electrodes did not harm speech-in-noise performance.

The effect of a coding strategy that removes temporally masked pulses on speech perception by cochlear implant users.

Speech recognition in noisy environments remains a challenge for cochlear implant (CI) recipients. Unwanted charge interactions between current pulses, both within and between electrode channels, are likely to impair performance. Here we investigate the effect of reducing the number of current pulses on speech perception. This was achieved by implementing a psychoacoustic temporal-masking model where current pulses in each channel were passed through a temporal integrator to identify and remove pulses that were less likely to be perceived by the recipient. The decision criterion of the temporal integrator was varied to control the percentage of pulses removed in each condition. In experiment 1, speech in quiet was processed with a standard Continuous Interleaved Sampling (CIS) strategy and with 25, 50 and 75% of pulses removed. In experiment 2, performance was measured for speech in noise with the CIS reference and with 50 and 75% of pulses removed. Speech intelligibility in quiet revealed no significant difference between reference and test conditions. For speech in noise, results showed a significant improvement of 2.4 dB when removing 50% of pulses and performance was not significantly different between the reference and when 75% of pulses were removed. Further, by reducing the overall amount of current pulses by 25, 50, and 75% but accounting for the increase in charge necessary to compensate for the decrease in loudness, estimated average power savings of 21.15, 40.95, and 63.45%, respectively, could be possible for this set of listeners. In conclusion, removing temporally masked pulses may improve speech perception in noise and result in substantial power savings.

Electro-haptic enhancement of speech-in-noise performance in cochlear implant users.

Cochlear implant (CI) users receive only limited sound information through their implant, which means that they struggle to understand speech in noisy environments. Recent work has suggested that combining the electrical signal from the CI with a haptic signal that provides crucial missing sound information ("electro-haptic stimulation"; EHS) could improve speech-in-noise performance. The aim of the current study was to test whether EHS could enhance speech-in-noise performance in CI users using: (1) a tactile signal derived using an algorithm that could be applied in real time, (2) a stimulation site appropriate for a real-world application, and (3) a tactile signal that could readily be produced by a compact, portable device. We measured speech intelligibility in multi-talker noise with and without vibro-tactile stimulation of the wrist in CI users, before and after a short training regime. No effect of EHS was found before training, but after training EHS was found to improve the number of words correctly identified by an average of 8.3%-points, with some users improving by more than 20%-points. Our approach could offer an inexpensive and non-invasive means of improving speech-in-noise performance in CI users.

Using recurrent neural networks to improve the perception of speech in non-stationary noise by people with cochlear implants.

Speech-in-noise perception is a major problem for users of cochlear implants (CIs), especially with non-stationary background noise. Noise-reduction algorithms have produced benefits but relied on a priori information about the target speaker and/or background noise. A recurrent neural network (RNN) algorithm was developed for enhancing speech in non-stationary noise and its benefits were evaluated for speech perception, using both objective measures and experiments with CI simulations and CI users. The RNN was trained using speech from many talkers mixed with multi-talker or traffic noise recordings. Its performance was evaluated using speech from an unseen talker mixed with different noise recordings of the same class, either babble or traffic noise. Objective measures indicated benefits of using a recurrent over a feed-forward architecture, and predicted better speech intelligibility with than without the processing. The experimental results showed significantly improved intelligibility of speech in babble noise but not in traffic noise. CI subjects rated the processed stimuli as significantly better in terms of speech distortions, noise intrusiveness, and overall quality than unprocessed stimuli for both babble and traffic noise. These results extend previous findings for CI users to mostly unseen acoustic conditions with non-stationary noise.

A Site-Selection Strategy Based on Polarity Sensitivity for Cochlear Implants: Effects on Spectro-Temporal Resolution and Speech Perception.

Thresholds of asymmetric pulses presented to cochlear implant (CI) listeners depend on polarity in a way that differs across subjects and electrodes. It has been suggested that lower thresholds for cathodic-dominant compared to anodic-dominant pulses reflect good local neural health. We evaluated the hypothesis that this polarity effect (PE) can be used in a site-selection strategy to improve speech perception and spectro-temporal resolution. Detection thresholds were measured in eight users of Advanced Bionics CIs for 80-pps, triphasic, monopolar pulse trains where the central high-amplitude phase was either anodic or cathodic. Two experimental MAPs were then generated for each subject by deactivating the five electrodes with either the highest or the lowest PE magnitudes (cathodic minus anodic threshold). Performance with the two experimental MAPs was evaluated using two spectro-temporal tests (Spectro-Temporal Ripple for Investigating Processor EffectivenesS (STRIPES; Archer-Boyd et al. in J Acoust Soc Am 144:2983-2997, 2018) and Spectral-Temporally Modulated Ripple Test (SMRT; Aronoff and Landsberger in J Acoust Soc Am 134:EL217-EL222, 2013)) and with speech recognition in quiet and in noise. Performance was also measured with an experimental MAP that used all electrodes, similar to the subjects' clinical MAP. The PE varied strongly across subjects and electrodes, with substantial magnitudes relative to the electrical dynamic range. There were no significant differences in performance between the three MAPs at group level, but there were significant effects at subject level-not all of which were in the hypothesized direction-consistent with previous reports of a large variability in CI users' performance and in the potential benefit of site-selection strategies. The STRIPES but not the SMRT test successfully predicted which strategy produced the best speech-in-noise performance on a subject-by-subject basis. The average PE across electrodes correlated significantly with subject age, duration of deafness, and speech perception scores, consistent with a relationship between PE and neural health. These findings motivate further investigations into site-specific measures of neural health and their application to CI processing strategies.

Comparison of effects on subjective intelligibility and quality of speech in babble for two algorithms: A deep recurrent neural network and spectral subtraction.

The effects on speech intelligibility and sound quality of two noise-reduction algorithms were compared: a deep recurrent neural network (RNN) and spectral subtraction (SS). The RNN was trained using sentences spoken by a large number of talkers with a variety of accents, presented in babble. Different talkers were used for testing. Participants with mild-to-moderate hearing loss were tested. Stimuli were given frequency-dependent linear amplification to compensate for the individual hearing losses. A paired-comparison procedure was used to compare all possible combinations of three conditions. The conditions were: speech in babble with no processing (NP) or processed using the RNN or SS. In each trial, the same sentence was played twice using two different conditions. The participants indicated which one was better and by how much in terms of speech intelligibility and (in separate blocks) sound quality. Processing using the RNN was significantly preferred over NP and over SS processing for both subjective intelligibility and sound quality, although the magnitude of the preferences was small. SS processing was not significantly preferred over NP for either subjective intelligibility or sound quality. Objective computational measures of speech intelligibility predicted better intelligibility for RNN than for SS or NP.

Vibro-Tactile Enhancement of Speech Intelligibility in Multi-talker Noise for Simulated Cochlear Implant Listening.

Many cochlear implant (CI) users achieve excellent speech understanding in acoustically quiet conditions but most perform poorly in the presence of background noise. An important contributor to this poor speech-in-noise performance is the limited transmission of low-frequency sound information through CIs. Recent work has suggested that tactile presentation of this low-frequency sound information could be used to improve speech-in-noise performance for CI users. Building on this work, we investigated whether vibro-tactile stimulation can improve speech intelligibility in multi-talker noise. The signal used for tactile stimulation was derived from the speech-in-noise using a computationally inexpensive algorithm. Eight normal-hearing participants listened to CI simulated speech-in-noise both with and without concurrent tactile stimulation of their fingertip. Participants' speech recognition performance was assessed before and after a training regime, which took place over 3 consecutive days and totaled around 30 min of exposure to CI-simulated speech-in-noise with concurrent tactile stimulation. Tactile stimulation was found to improve the intelligibility of speech in multi-talker noise, and this improvement was found to increase in size after training. Presentation of such tactile stimulation could be achieved by a compact, portable device and offer an inexpensive and noninvasive means for improving speech-in-noise performance in CI users.

Use of a Deep Recurrent Neural Network to Reduce Wind Noise: Effects on Judged Speech Intelligibility and Sound Quality.

Despite great advances in hearing-aid technology, users still experience problems with noise in windy environments. The potential benefits of using a deep recurrent neural network (RNN) for reducing wind noise were assessed. The RNN was trained using recordings of the output of the two microphones of a behind-the-ear hearing aid in response to male and female speech at various azimuths in the presence of noise produced by wind from various azimuths with a velocity of 3 m/s, using the "clean" speech as a reference. A paired-comparison procedure was used to compare all possible combinations of three conditions for subjective intelligibility and for sound quality or comfort. The conditions were unprocessed noisy speech, noisy speech processed using the RNN, and noisy speech that was high-pass filtered (which also reduced wind noise). Eighteen native English-speaking participants were tested, nine with normal hearing and nine with mild-to-moderate hearing impairment. Frequency-dependent linear amplification was provided for the latter. Processing using the RNN was significantly preferred over no processing by both subject groups for both subjective intelligibility and sound quality, although the magnitude of the preferences was small. High-pass filtering (HPF) was not significantly preferred over no processing. Although RNN was significantly preferred over HPF only for sound quality for the hearing-impaired participants, for the results as a whole, there was a preference for RNN over HPF. Overall, the results suggest that reduction of wind noise using an RNN is possible and might have beneficial effects when used in hearing aids.

Tolerable delay for speech production and perception: effects of hearing ability and experience with hearing aids.

OBJECTIVE: Processing delay is one of the important factors that limit the development of novel algorithms for hearing devices. In this study, both normal-hearing listeners and listeners with hearing loss were tested for their tolerance of processing delay up to 50 ms using a real-time setup for own-voice and external-voice conditions based on linear processing to avoid confounding effects of time-dependent gain. DESIGN: Participants rated their perceived subjective annoyance for each condition on a 7-point Likert scale. STUDY SAMPLE: Twenty normal-hearing participants and twenty participants with a range of mild to moderate hearing losses. RESULTS: Delay tolerance was significantly greater for the participants with hearing loss in two out of three voice conditions. The average slopes of annoyance ratings were negatively correlated with the degree of hearing loss across participants. A small trend of higher tolerance of delay by experienced users of hearing aids in comparison to new users was not significant. CONCLUSION: The increased tolerance of processing delay for speech production and perception with hearing loss and reduced sensitivity to changes in delay with stronger hearing loss may be beneficial for novel algorithms for hearing devices but the setup used in this study differed from commercial hearing aids.

Auditory inspired machine learning techniques can improve speech intelligibility and quality for hearing-impaired listeners.

Machine-learning based approaches to speech enhancement have recently shown great promise for improving speech intelligibility for hearing-impaired listeners. Here, the performance of three machine-learning algorithms and one classical algorithm, Wiener filtering, was compared. Two algorithms based on neural networks were examined, one using a previously reported feature set and one using a feature set derived from an auditory model. The third machine-learning approach was a dictionary-based sparse-coding algorithm. Speech intelligibility and quality scores were obtained for participants with mild-to-moderate hearing impairments listening to sentences in speech-shaped noise and multi-talker babble following processing with the algorithms. Intelligibility and quality scores were significantly improved by each of the three machine-learning approaches, but not by the classical approach. The largest improvements for both speech intelligibility and quality were found by implementing a neural network using the feature set based on auditory modeling. Furthermore, neural network based techniques appeared more promising than dictionary-based, sparse coding in terms of performance and ease of implementation.