Tobias Goehring

The paper investigates how neural encoding of speech varies with acoustic degradation and language, finding that flexible encoding strategies predict comprehension of degraded speech.

The paper decodes speech acoustics from EEG using extended features beyond amplitude envelopes, showing linear models perform comparably to non-linear ones and some features correlate with speech intelligibility.

This study decodes speech envelope neural activity using EEG in hearing listeners, finding that intelligible speech improves decoding accuracy despite no effect of spectral resolution.