An Algorithm for Intelligibility Prediction of Time–Frequency Weighted Noisy Speech

TLDR

An objective machine‑driven intelligibility measure that correlates highly with speech intelligibility is sought in noise‑reduction algorithm development. The study presents STOI, a short‑time objective intelligibility measure, to reduce testing time and to guide improvements in noisy speech intelligibility. STOI computes intelligibility from 386‑ms short‑time segments and is implemented in freely available MATLAB code. STOI outperforms five other models in correlation with speech intelligibility, confirming the advantage of 386‑ms segment lengths.

Abstract

In the development process of noise-reduction algorithms, an objective machine-driven intelligibility measure which shows high correlation with speech intelligibility is of great interest. Besides reducing time and costs compared to real listening experiments, an objective intelligibility measure could also help provide answers on how to improve the intelligibility of noisy unprocessed speech. In this paper, a short-time objective intelligibility measure (STOI) is presented, which shows high correlation with the intelligibility of noisy and time-frequency weighted noisy speech (e.g., resulting from noise reduction) of three different listening experiments. In general, STOI showed better correlation with speech intelligibility compared to five other reference objective intelligibility models. In contrast to other conventional intelligibility models which tend to rely on global statistics across entire sentences, STOI is based on shorter time segments (386 ms). Experiments indeed show that it is beneficial to take segment lengths of this order into account. In addition, a free Matlab implementation is provided.