A computational model that simulates loudness perception experiments using a peripheral auditory model for electric and acoustic stimulation (PAMEAS).
This study introduces a new method for designing optimized multipolar stimulation strategies, offering improved auditory perception with potentially reduced energy consumption.
We develop methods that (1) enable accurate design of auditory filterbanks based on desired specifications for peak frequencies, bandwidths, and delays and (2) allow for very systematically studying the effect of varying filter characteristics on perceptual tasks and on improving speech and hearing technologies.
This project proposes a hearing aid framework based on Artificial Neural Network (ANN) and Cascade of Asymmetric Resonators with Fast-Acting Compression (CARFAC) model.
In this paper, it has been shown that a novel electrical stimulation strategy that is based on healthy cochlea, can produce neurograms that resemble the simulated neural activity of a healthy cochlea more that those produced by a current algorithm used in cochlear implants (ACE).
We show statistical analysis of the auditory brainstem response (ABR) to estimate the latent signal and noise levels, using a regression approach, and show the empirical errors as a function of these levels and the number of trials.
We use Bayesian modeling to connect audiology tests such as speech in noise (SPIN) and not only predict missing data, but also reason about the quality of our models.
Here we explore how hearing loss affects attentional subprocesses such as target selection and distractor suppression, using EEG to examine alpha modulation patterns in normal-hearing young and elderly individuals, as well as elderly individuals with hearing loss.
Pitch contour perception in speech and music at accelerated speed in hearing-impaired seniors.
This work introduces a new real-time usable and non-intrusive version of the Binaural Speech Intelligibility Model (BSIM).
Spatial processing disorder, a long-term consequence of otitis media early in life, is an extremely common cause of speech in noise deficits that occur despite hearing thresholds being normal.
We modeled normal and impaired auditory perception with an artificial neural network trained to localize and recognize sounds from simulated auditory nerve input.
This study presents a comprehensive computational modeling framework of the peripheral auditory system, integrating 3D volume geometries with numerical simulation to investigate the mechanisms of electrically and acoustically stimulated cochlear implants, with potential applications in clinical fitting, neural health assessment, and perceptual tasks for cochlear implant patients.
Computational model that includes customizability of biophysical parameters for simulating electrocochleography responses in cochlear implant users.
We used modelling to investigate whether phenomena such as refractoriness and facilitation as early as the auditory nerve are sufficient to create a neural distortion response.
This workshop aims to demonstrate how openly accessible auditory models can help us simulate auditory processing for normal-hearing and hearing-impaired listeners, and how they can be used to develop optimal sound-processing strategies for real-world applications such as hearing aids.
In this presentation we explore Deep ACE, an end-to-end speech-denoising sound coding strategy for cochlear implants. Its performance is evaluated using vSAMII, an objective metric based on mutual information that incorporates a physiological model of the human auditory system.
We propose an adaptive leaky-integrate and firing probability (aLIFP) model of an electrically stimulated auditory nerve fiber, which outputs the spike probability distribution over time and accounts for refractoriness, facilitation, accommodation, and long-term adaptation.
Acoustic response of ear canal combined between ear simulator and acoustic coupled cavity by numerical simulation computation and actual measurement.
Using models of acoustic and electric hearing, we were able to visualize unintentional cues in the spectral ripple test and spectrally modulated ripple test.
This presentation will introduce a framework used for developing and deploying cloud-based machine learning models in the context of a project aimed at optimizing manufacturing of custom (in-the-ear style) hearing aids by improved detection of electroacoustic test failures due to feedback.
A computational model of the auditory periphery applicable to round window stimulation.
A novel deep learning model with a convolutional multi-task learning architecture to effectively predict noise-induced permanent threshold shifts.
This study investigates using deep neural networks to provide hearing loss compensation for people with hearing loss, showing promising results in speech intelligibility and perceived quality.
We extend auditory filters to have rational exponents – and derive corresponding implementations, in order to achieve more flexible filter behavior and control the filter characteristics on a continuum not accessible by integer order filters.
Effect of unilateral and bilateral cochlear implants on speech recognition performance in spatialized noise can be simulated using a channel vocoder with dense tonal carriers.