Harmonic Beamformers for Non-Intrusive Speech Intelligibility Prediction.

In recent years, research into objective speech intelligibility measures has gained increased interest as a tool to optimize speech enhancement algorithms. While most intelligibility measures are intrusive, i.e., they require a clean reference signal, this is rarely available in real-time applications. This paper proposes two non-intrusive intelligibility measures, which allow using the intrusive short-time objective intelligibility (STOI) measure without requiring access to the clean signal. Instead, a reference signal is obtained from the degraded signal using either a fixed or an adaptive harmonic spatial filter. This reference signal is then used as input to STOI. The experimental results show a high correlation between both proposed non-intrusive speech intelligibility measures and the original intrusively computed STOI scores.