A real-time wavelet-based algorithm for improving speech intelligibility

Abstract

A wavelet-based algorithm to improve speech intelligibility is reported. The speech signal is split into frequency sub-bands via a multi-level discrete wavelet transform. Various gains are applied to the sub-band signals before they are recombined to form a modified version of the speech. Dynamic range compression then follows to control the peak amplitude. The sub-band gains are adjusted while keeping the overall signal energy unchanged, and the speech intelligibility under simulated hearing loss conditions and various background interference is enhanced and evaluated objectively and quantitatively using Google Speech-to-Text transcription. For English and Chinese noise-free speech, overall intelligibility is improved, and the transcription accuracy can increase by over 80 percentage points by reallocating the spectral energy toward the mid-frequency sub-bands, effectively increasing the consonant-vowel intensity ratio. This is reasonable since the consonants are relatively weak and of short duration, and are therefore the most likely to become indistinguishable in the presence of background noise or high frequency hearing impairment. For speech already corrupted by noise, improving intelligibility is challenging but still realizable. The proposed algorithm is implementable in real-time and comparatively simpler than previous algorithms. Potential applications include speech transmission, hearing aids, machine listening, and a better understanding of speech intelligibility.