Modulation masking and fine structure shape neural envelope coding to predict speech intelligibility across diverse...

Area of Science:

• Neuroscience
• Auditory Perception
• Speech Processing

Background:

• Understanding how acoustic scene features influence neural processing of attended speech is crucial for communication.
• The role of background sounds in modulating neural encoding of speech envelopes and its impact on intelligibility remains unclear.

Purpose of the Study:

• To investigate how scene acoustics, specifically background sounds, affect neural processing of attended speech.
• To determine the relationship between neural encoding of speech envelopes and speech intelligibility in realistic acoustic conditions.

Main Methods:

• Combined human electroencephalography (EEG) with simultaneous speech intelligibility measurements.
• Utilized high-resolution vocoding to manipulate peripheral envelopes and assess the role of temporal fine structure (TFS).

Main Results:

• Neural envelope-domain signal-to-noise ratio in target speech encoding predicted intelligibility across various realistic listening conditions.
• Evidence for modulation masking was found, where masker modulations shape neural encoding.
• Speech envelope coding fidelity depended on both cochlear-conveyed envelopes and TFS, which is critical for scene segregation.

Conclusions:

• Temporal coherence of sound elements influences scene analysis and attention to target sounds.
• Findings provide neurophysiological evidence for modulation masking and highlight the importance of TFS in speech perception.
• Results can inform speech intelligibility models and technologies for real-world communication enhancement.