Effectiveness of Two-Talker Maskers That Differ in Talker Congruity and Perceptual Similarity to the Target Speech.
Effectiveness of Two-Talker Maskers That Differ in Talker Congruity and Perceptual Similarity to the Target Speech.
复制标题
DOI：
10.1177/2331216517709385
复制
发表时间：
2017-01
影响因子：
2.7
通讯作者：
Bowdrie K
中科院分区：
文献类型：
--
作者：
Calandruccio L;Buss E;Bowdrie K
先前的研究表明，当掩蔽者由两个相互竞争的说话者组成时，掩蔽句的识别效果特别差，这一发现归因于信息掩蔽。当目标和掩蔽者说话者在感知上相似时，信息掩蔽往往是最大的。掩蔽的减少已经被观察到了广泛的目标和掩蔽者的差异，包括语言：与相同的语言相比，当目标和掩蔽者用不同的语言说话时，表现更好。本研究评估了听力正常的成年人的句子识别在两个说话者掩蔽作为目标和两个掩蔽流之间的感知相似性的函数。目标是英国人，面具人由英语，时间逆转的英语或荷兰语组成。这三种类型的掩蔽者在信息掩蔽方面各不相同。两个谈话者掩蔽物内的两个谈话者是一致的（例如，都是英语）或不一致（例如，一个英国人，一个荷兰人）。正如所预测的那样，平均表现是更糟糕的一致的英语面具比一致的时间颠倒的英语或一致的荷兰面具。不一致的两个谈话者的面具，只有一个英语面具流，只是稍微低于一致的英语面具的有效性。这一结果表明，两个谈话者掩蔽效果主要是由一个掩蔽流，是最感知相似的目标。语音识别在一个单一的谈话者掩蔽之间的英语，荷兰语，和时间反转的英语掩蔽类型只有轻微的不同，这表明感知相似性可能是更重要的两个谈话者比一个谈话者掩蔽。
Previous work has shown that masked-sentence recognition is particularly poor when the masker is composed of two competing talkers, a finding that is attributed to informational masking. Informational masking tends to be largest when the target and masker talkers are perceptually similar. Reductions in masking have been observed for a wide range of target and masker differences, including language: Performance is better when the target and masker talkers speak in different languages, compared with the same language. The present study evaluated normal-hearing adults’ sentence recognition in a two-talker masker as a function of the perceptual similarity between the target and each of the two masker streams. The target was English, and the maskers were composed of English, time-reversed English, or Dutch. These three masker types are known to vary in the informational masking they exert. The two talkers within the two-talker maskers were either congruent (e.g., both English) or incongruent (e.g., one English, one Dutch). As predicted, mean performance was worse for the congruent English masker than the congruent time-reversed English or congruent Dutch maskers. Incongruent two-talker maskers, with just one English masker stream, were only modestly less effective than the congruent English masker. This result indicates that two-talker masker effectiveness was determined predominantly by the one masker stream that was most perceptually similar to the target. Speech recognition in a single-talker masker differed only marginally between the English, Dutch, and time-reversed English masker types, suggesting that perceptual similarity may be more critical in a two-talker than a one-talker masker.