Answers are grounded in this paper and cite their sources.
Explores speech intelligibility with across-frequency coherence in varied age and hearing status, suggesting pathways for improved models.
Older normal-hearing (NH) and hearing-impaired (HI) listeners often struggle understanding speech in adverse conditions. Identifying the underlying suprathreshold deficits contributing to these challenges remains difficult. Motivated by auditory-modeling studies highlighting the role of across-frequency coherence in speech-intelligibility (SI) prediction, we used a “distortion-sensitivity” approach to explore listeners’ perception of across-frequency phase coherence as a potential suprathreshold factor influencing SI. We manipulated across-frequency coherence by adding random phase components to each frequency channel of speech-and-noise mixtures presented at 5-dB signal-to-noise ratio. SI was measured as a function of phase-distribution spread. Additionally, we obtained just-noticeable-differences (JNDs) for phase-jittered complex tones and speech reception thresholds (SRTs) for undistorted speech in the presence of stationary, fluctuating, and speech-like maskers. Data from 10 young NH, 10 older NH, and 9 older HI listeners showed highest JNDs for the HI group, followed by the older NH and young NH listeners, suggesting reduced sensitivity to across-frequency coherence with age and hearing loss. Despite group differences in JNDs, no significant differences in SI performance for jittered speech were observed across groups. However, SRTs for undistorted speech exhibited trends consistent with the JND patterns. This work sets the basis for improving individualized SI models by incorporating phase-distortion sensitivity.
No takes yet. Share an insight, caveat, or question.
Relaño-Iborra et al. (2025) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: