Sound masking efficiency of echo-time-reversed masker

This paper investigates the masking efficiency of a new sound masker named echo-time-reversed masker. The masker is obtained by generating an acoustic echo from the target speech and then processing it with time reversal method. The masking efficiency of the echo-time-reversed masker is evaluated by two objective intelligibility measurements, i. e. the perceptual evaluation of speech quality (PESQ) measure and the modified normalized covariance metric (NCM) measure. Results show that the intelligibility of the masked speech decreases as the target-to-masker ratio (TMR) decreases, or the frame length increases. Besides, the masking efficiency of echo-time-reversed masker and that of time-reversed masker are compared as well. Results show that with the same target-to-masker ratio (TMR), the echo-time-reversed masker is much more efficient in reducing speech intelligibility.