Speech quality assessment is indispensable to properly design a speech enhancement algorithm. The perceptual evaluation of speech quality (PESQ) is frequently employed as an objective speech distortion measure. The PESQ is a methodology for estimating subjective assessment of speech quality assuming a slight distortion caused by speech codecs for telephony systems. In case of noise reduction, however, a degree of speech distortion is heavier than those caused by the speech codecs. In this paper applicability of the PESQ is investigated for noisy and noise-reduced speech signals under severe noisy conditions. A relationship between PESQ scores and subjective mean opinion scores reveals that the PESQ can be applicable for heavily distorted speech only under non-stationary noisy conditions.
(See document for exact affiliation information.)
Click to purchase paper as a non-member or login as an AES member. If your company or school subscribes to the E-Library then switch to the institutional version. If you are not an AES member Join the AES. If you need to check your member status, login to the Member Portal.
