Title: “Binaural benefit in speech intelligibility with spatially separated speech maskers”
Dr. H. Steven Colburn, BU BME (Advisor)
Dr. David C. Mountain, BU BME (Chair)
Dr. Kamal Sen, BU BME
Dr. Gerald Kidd Jr, BU Sargent College
Dr. Virginia Best, National Acoustic Labs of Australia
The goal of this work is to better understand why some listeners have an easier time than others listening to a target talker amid spatially-distributed interfering talkers. Speech reception thresholds (SRTs) of young normal-hearing and young hearing-impaired listeners were measured, with the target in-front, and two maskers arranged in the following spatial configurations: 1) “colocated”, both in-front, 2) “symmetric”, maskers at ± 60⁰ azimuth, and 3) “anti-symmetric”, both at either +60⁰, or -60⁰ azimuth. SRTs were measured both binaurally and monaurally, with monaural stimuli presented to the “better-ear.” Binaural benefits were calculated as the decrease in threshold when the second ear was added to the better-ear. In addition, sensitivity to interaural time differences, interaural level differences and interaural cross-correlation were measured in the same listeners, using narrow-band noises centered at a low frequency (500 Hz) and at a high frequency (either 2 kHz or 4 kHz). Finally, SRTs were predicted using the short-time Equalization-Cancellation (EC) model, in which two types of parameters were varied: first, the time- and intensity-jitters at the inputs of the binaural processing mechanism (which affects interaural discrimination abilities) and second, the temporal window in the binaural processing mechanism (which affects the ability to benefit from short-time binaural advantages).
Large inter-subject differences were found among both listener groups in binaural benefits in both the symmetric and anti-symmetric cases, and also in interaural difference sensitivity thresholds. Symmetric binaural benefit did not correlate with interaural difference sensitivity, as measured within either listener group. Changing the amount of jitter had a relatively small effect on symmetric predicted SRTs, while changing the EC- window length had a larger effect. Anti-symmetric binaural benefit correlated with thresholds measured in many of the interaural difference sensitivity tasks, both within the normal-hearing group and among a subset of hearing-impaired listeners. Changing the amount of jitter had a large effect on anti-symmetric predicted SRTs, while varying the length of the EC window had a small effect. These results suggest some key differences between the symmetric and anti-symmetric cases. Both interaural difference sensitivity and a mechanism analogous to the length of the EC window might be factors in the symmetric case, while interaural difference sensitivity is a key factor in the anti-symmetric case.