An official website of the United States government
We propose a new adaptive feedback cancellation (AFC) system in hearing aids (HAs) based on a well-posed optimization criterion that jointly considers both decorrelation of the signals and sparsity of the underlying channel. We show that the least squares criterion on subband errors regularized by a p-norm-like diversity measure can be used to simultaneously decorrelate the speech signals and exploit sparsity of the acoustic feedback path impulse response. Compared with traditional subband adaptive filters that are not appropriate for incorporating sparsity due to shorter sub-filters, our proposed framework is suitable for promoting sparse characteristics, as the update rule utilizing subband information actually operates in the fullband. Simulation results show that the normalized misalignment, added stable gain, and other objective metrics of the AFC are significantly improved by choosing a proper sparsity promoting factor and a suitable number of subbands. More importantly, the results indicate that the benefits of subband decomposition and sparsity promoting are complementary and additive for AFC in HAs.
more »
« less
On Mitigating Acoustic Feedback in Hearing Aids with Frequency Warping by All-Pass Networks
Acoustic feedback control continues to be a challenging prob- lem due to the emerging form factors in advanced hearing aids (HAs) and hearables. In this paper, we present a novel use of well-known all-pass filters in a network to perform frequency warping that we call “freping.” Freping helps in breaking the Nyquist stability criterion and improves adaptive feedback can- cellation (AFC). Based on informal subjective assessments, dis- tortions due to freping are fairly benign. While common ob- jective metrics like the perceptual evaluation of speech quality (PESQ) and the hearing-aid speech quality index (HASQI) may not adequately capture distortions due to freping and acoustic feedback artifacts from a perceptual perspective, they are still instructive in assessing the proposed method. We demonstrate quality improvements with freping for a basic AFC (PESQ: 2.56 to 3.52 and HASQI: 0.65 to 0.78) at a gain setting of 20; and an advanced AFC (PESQ: 2.75 to 3.17 and HASQI: 0.66 to 0.73) for a gain of 30. From our investigations, freping provides larger improvement for basic AFC, but still improves overall system performance for many AFC approaches.
more »
« less
- Award ID(s):
- 1838897
- PAR ID:
- 10351868
- Date Published:
- Journal Name:
- Proceedings Article published 15 Sep 2019 in Interspeech 2019
- Page Range / eLocation ID:
- 4245 to 4249
- Format(s):
- Medium: X
- Sponsoring Org:
- National Science Foundation
More Like this
-
-
Speech sounds exist in a complex acoustic–phonetic space, and listeners vary in the extent to which they are sensitive to variability within the speech sound category (“gradience”) and the degree to which they show stable, consistent responses to phonetic stimuli. Here, we investigate the hypothesis that individual differences in the perception of the sound categories of one's language may aid speech-in-noise performance across the adult lifespan. Declines in speech-in-noise performance are well documented in healthy aging, and are, unsurprisingly, associated with differences in hearing ability. Nonetheless, hearing status and age are incomplete predictors of speech-in-noise performance, and long-standing research suggests that this ability draws on more complex cognitive and perceptual factors. In this study, a group of adults ranging in age from 18 to 67 years performed online assessments designed to measure phonetic category sensitivity, questionnaires querying recent noise exposure history and demographic factors, and crucially, a test of speech-in-noise perception. Results show that individual differences in the perception of two consonant contrasts significantly predict speech-in-noise performance, even after accounting for age and recent noise exposure history. This finding supports the hypothesis that individual differences in sensitivity to phonetic categories mediates speech perception in challenging listening situations.more » « less
-
Auditory masking—the interference of the encoding and processing of an acoustic stimulus imposed by one or more competing stimuli—is nearly omnipresent in daily life and presents a critical barrier to many listeners, including people with hearing loss, users of hearing aids and cochlear implants, and people with auditory processing disorders. The perceptual aspects of masking have been actively studied for several decades, and particular emphasis has been placed on masking of speech by other speech sounds. The neural effects of such masking, especially at the subcortical level, have been much less studied, in large part due to the technical limitations of making such measurements. Recent work has allowed estimation of the auditory brainstem response (ABR), whose characteristic waves are linked to specific subcortical areas, to naturalistic speech. In this study, we used those techniques to measure the encoding of speech stimuli that were masked by one or more simultaneous other speech stimuli. We presented listeners with simultaneous speech from one, two, three, or five simultaneous talkers, corresponding to a range of signal-to-noise ratios (clean, 0, −3, and −6 dB), and derived the ABR to each talker in the mixture. Each talker in a mixture was treated in turn as a target sound masked by other talkers, making the response quicker to acquire. We found consistently across listeners that ABR Wave V amplitudes decreased and latencies increased as the number of competing talkers increased.more » « less
-
In speech perception, when a primary acoustic cue (e.g., VOT) is ambiguous, listeners may increase the weight of a secondary cue (e.g., F0). In experiment 1, we compared the cue-weighting adjustment strategies across younger and older normal-hearing adults with a distributional learning paradigm. Two groups of native English listeners were exposed to voicing contrasts that were ambiguous in either VOT or F0. Additionally, listeners may access lexical information to help resolve the ambiguity in the acoustic signal. Older listeners have been reported to use lexical information to a greater degree than younger listeners. In experiment 2, using a lexically guided learning paradigm, we tested if younger and older adults differ in their use of lexical information when learning to interpret ambiguous acoustic tokens. There were four types of exposure, in which stimuli differed in lexical status (day-*tay; *doy-toy) and the acoustic ambiguity involved either only VOT or both VOT and F0. Preliminary results from younger normal-hearing, listeners showed significant speech adaptation effects, with a significant change in cue weights in distributional learning and salient lexical bias in lexically guided learning. More data will be collected from older adults to assess the extent of perceptual learning relative to younger adults.more » « less
-
This dissertation compared speech perception across younger and older normal-hearing adults. We ask four research questions to assess acoustic cue weighting and the role of contextual information (lexical information and speaking rate) in speech perception. Experiment 1 tested for age-related changes in cue-weighting. The absolute weights showed that older listeners relied on both VOT and F0 more than younger listeners, and some listeners’ reliance on VOT correlated with inhibitory control when perceiving the /d/-/t/ contrast. The relative weights suggested that older listeners relied on VOT less and F0 more than younger listeners. Experiment 2 tested whether younger and older listeners used different cue-weighting adjustment strategies in distributional learning. Both older and younger listeners adjusted their reliance on acoustic cues when the primary acoustic cue (VOT) became ambiguous. Older listeners adjusted F0 to a greater degree than younger listeners, while younger listeners adjusted VOT more than older listeners. With a lexically-guided learning paradigm, Experiment 3 explored if younger and older adults differed in their use of lexical information when learning to map acoustic tokens that were ambiguous. Older and younger listeners utilized the lexical context to the same extent. In Experiment 4, the contextual effect of speaking rate was examined by embedding voicing contrasts in short and long syllables and presenting these syllables to younger and older listeners. Older and younger listeners compensated for variation in speaking rate in a similar manner as younger listeners. The findings in perceptual learning demonstrate perceptual flexibility among normal-hearing older listeners, despite an assumed decline in temporal processing.more » « less
- Free Publicly Accessible Full Text
-
Accepted Manuscript
- Conference Paper:
- https://doi.org/10.21437/INTERSPEECH.2019-3195
