Speech intelligibility prediction based on modulation frequency-selective processing

Relaño-Iborra, Helia; Dau, Torsten

doi:10.1016/j.heares.2022.108610

Cited by 12 publications

(6 citation statements)

References 114 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This frequency-selective characteristic of AM processing has been modelled using the concept of a modulation filterbank, based on the idea that AM fluctuations are decomposed through an array of relatively broad bandpass modulation filters with a constant quality (Q) factor of approximately 1-2 (e.g., Dau et al ., 1997a, 1999; Ewert and Dau, 2000). Computational modelling studies have successfully applied the modulation filterbank concept to simulate data from various experimental paradigms, including simultaneous and non-simultaneous spectral and temporal signal detection and masking conditions (Dau et al ., 1997a, 1997b, 1999; Verhey et al ., 1999; Ewert and Dau, 2000; Ewert et al ., 2002; Piechowiak et al ., 2007; Jepsen et al ., 2008; Jepsen and Dau, 2011; King et al ., 2019), sound texture perception (McDermott and Simoncelli, 2011; McDermott et al ., 2013; McWalter and Dau, 2015, 2017), auditory stream segregation (Elhilali et al ., 2009; Christiansen et al ., 2014), and speech intelligibility (Jørgensen and Dau, 2011; Jørgensen et al, 2013; Relaño-Iborra et al, 2016, 2019; Zaar and Dau, 2017; Zaar et al, 2017; Steinmetzger et al, 2019; Zaar and Carney, 2022; for a review, see Relaño-Iborra and Dau, 2022). Furthermore, the modulation filterbank is conceptually consistent with the temporal dimension of a ‘two-dimensional’ spectro-temporal modulation filterbank, inspired by neural responses to spectro-temporally varying stimuli in the auditory cortex of ferrets (Kowalski et al ., 1996; Depireux et al ., 2001) and supported by data from perceptual learning and masking conditions (Sabin et al ., 2012; Oetjen and Verhey, 2015, 2017; Conroy et al ., 2022), as well as models of speech intelligibility (Elhilali et al ., 2003; Chi et al ., 2005; Zilany and Bruce, 2007; Chabot-Leclerc et al ., 2014).…”

Section: Introductionmentioning

confidence: 99%

Disentangling the effects of hearing loss and age on amplitude modulation frequency selectivity

Regev

Relaño-Iborra

Zaar

et al. 2023

Preprint

Self Cite

View full text Add to dashboard Cite

The processing and perception of amplitude modulations (AMs) in the auditory system reflect a frequency-selective process, often described as a modulation filterbank. Previous studies on perceptual AM masking reported similar results for older listeners with hearing impairment (HI) and young listeners with normal hearing (NH), suggesting no effects of age nor hearing loss on AM frequency selectivity. However, recent evidence has shown that age, independently of hearing loss, is detrimental to AM frequency selectivity. Hence, the present study aimed to disentangle the effects of hearing loss and age. A simultaneous AM masking paradigm was employed, utilizing a sinusoidal carrier at 2.8 kHz, narrow-band noise modulation maskers, and target modulation frequencies of 4, 16, 64, and 128 Hz. The results obtained from older (n=10, 63-77 years) and young HI listeners (n=3, 24-30 years) were compared to data from young and older NH listeners. Notably, the HI listeners generally exhibited lower (unmasked) AM detection thresholds and greater AM frequency selectivity compared to their NH counterparts of similar age. These findings suggest that age negatively affects AM frequency selectivity in both NH and HI listeners, while hearing loss improves AM detection and AM selectivity, likely due to the loss of peripheral compression.

show abstract

Section: Introductionmentioning

confidence: 99%

Disentangling the effects of hearing loss and age on amplitude modulation frequency selectivity

Regev

Relaño-Iborra

Zaar

et al. 2023

Preprint

Self Cite

View full text Add to dashboard Cite

show abstract

“…Auditory models [1][2][3][4] provide insights into the inner workings of the auditory system and can be a useful tool when, e.g., analyzing novel audio-processing techniques and hearing-aid algorithms. More importantly, models provide a valuable framework to test hypotheses regarding potential effects of hearing deficits in the auditory system on relevant outcome measures.…”

Section: Evaluating An Auditory Model As Predictor Of Speech Understa...mentioning

confidence: 99%

Evaluating an Auditory Model as Predictor of Speech Understanding in Hearing-Impaired Listeners

Relaño-Iborra,

Zaar,

Dau

2024

Proceedings of the 10th Convention of the European Acoustics Association Forum Acusticum 2023

View full text Add to dashboard Cite

Speech intelligibility (SI) models that represent effects of hearing impairment (HI) in their processing stages may help understand the link between clinical measures of auditory dysfunction and daily-life challenges with speech-in-noise understanding experienced by listeners. Here, we present a thorough evaluation of a well-established normal-hearing (NH) SI model, the speech-based computational auditory signal processing and perception model [1], as a predictor of SI performance in HI listeners. By modelling three previously published datasets, we evaluated the model's predictive power of (i) the role of audibility in unaided speech reception thresholds, (ii) the masking release obtained in fluctuating (relative to stationary) maskers, and (iii) aided performance in conditions where amplification was provided. The model was evaluated both at the listener group level to assess differences between NH and HI populations, and as a predictor of individual listener performance within the HI population. For comparison, clinical estimates used to fit the model were also analyzed as individual SI predictors. We show that the current model accurately captures effects of audibility and represents the loss of masking release observed in the HI listeners. However, the model is still limited in accounting for data associated with supra-threshold auditory deficits.

show abstract

“…A more realistic implementation of this stage in the overall CASP model is thus hypothesised to improve model predictions, in particular with regard to predictions of data from listeners with a hearing impairment. The large variability observed in data across listeners with hearing impairment, especially in speech-related tasks, is still largely underestimated by current modelling studies [6]. The present study aimed to validate the revised CASP model implementation in psychoacoustic conditions for listeners with normal hearing.…”

Section: Introductionmentioning

confidence: 99%

Evaluation of the revised computational model of auditory signal processing and perception

Paulick,

Relaño-Iborra,

Dau

2022

Proceedings of the 10th Convention of the European Acoustics Association Forum Acusticum 2023

View full text Add to dashboard Cite

A revised version of the computational auditory signal processing and perception model (CASP, [1]) is proposed. CASP is a functional model designed for the prediction of data from various psychoacoustic conditions, including intensity discrimination, spectral and temporal masking and amplitude modulation detection and masking. The revised version of CASP proposed here introduces two main changes to the original model: (i) a more realistic, nonlinear inner hair cell stage replacing the simplified linear processing stage of the original model and (ii) an updated back-end processing that captures off-frequency listening through an automatic channel selection algorithm. The effects of these two modifications in the model are analysed in various signal detection and masking conditions, including intensity discrimination, forward masking as well as modulation detection using narrowband noise carriers. Model predictions are compared across model configurations and examined in terms of the internal representations of the signals at different stages of auditory processing within the model. The potential implications of the revised signal processing in the model on predictions of effects of sensorineural hearing loss are discussed.

show abstract

Speech intelligibility prediction based on modulation frequency-selective processing

Cited by 12 publications

References 114 publications

Disentangling the effects of hearing loss and age on amplitude modulation frequency selectivity

Disentangling the effects of hearing loss and age on amplitude modulation frequency selectivity

Evaluating an Auditory Model as Predictor of Speech Understanding in Hearing-Impaired Listeners

Evaluation of the revised computational model of auditory signal processing and perception

Contact Info

Product

Resources

About