论文部分内容阅读
通过心理声学实验研究了来自不同方向具有不同信噪比的两种干扰声条件下,母语为汉语的听者对英语的空间去掩蔽现象。在消声室指定位置布放扬声器,发出目标声和干扰声,通过听者对目标声进行听音识别,得到听者识别的正确率。实验结果显示:只在正前方播放目标语音时,识别正确率大于99%,当目标和干扰语音都位于听者正前方时,正确率为57%;当目标和干扰语音随机位于士60°时,正确率为96%;特别地,当目标语音和干扰信号都位于听者正前方时,若干扰为噪声,随着信噪比从0 dB降低到-12 dB,正确率从96%降低到34%,而当干扰为语音时,随着信噪比从0 dB降低到-12 dB,正确率先是下降,随后有平均幅度为27%的明显上升,在此之后又是下降的趋势;当噪声干扰和语音干扰位于60°时,随着信噪比从-4 dB降低到-16 dB,正确率分别从99%降低到80%和从98%降低到91%。研究表明:空间分离对于母语为汉语的听者的英语语音可懂度有明显增益;大多数情况下英语语音的正确率都随着信噪比的降低而下降。这和对母语为其他语言的相关研究结论一致。
Through the psychoacoustic experiments, we study the spatial de-masking of English speakers from Chinese listeners under two kinds of interference noise with different signal-to-noise ratios in different directions. Speakers in the anechoic chamber at the designated location, the target sound and interference sound, through the listener on the target sound for listening recognition, the listener to recognize the correct rate. The experimental results show that the recognition accuracy is greater than 99% when the target speech is played directly in front of the listener, and the correct rate is 57% when both the target speech and the interfering speech are located directly in front of the listener. When the target and the interfering speech are randomly located at ± 60 ° , With a correct rate of 96%. In particular, when both the target speech and the interference signal are directly in front of the listener, if the interference is noise, the accuracy decreases from 96% as the signal-to-noise ratio decreases from 0 dB to -12 dB 34%. When the signal to noise ratio decreases from 0 dB to -12 dB, the correct rate first decreases and then the average increase rate of 27% increases sharply, and then decreases again. When Noise and speech interference At 60 °, the correct rate was reduced from 99% to 80% and from 98% to 91%, respectively, as the signal-to-noise ratio decreased from -4 dB to -16 dB. The results show that: Spatial separation has a significant gain for English speech intelligibility of native speakers; in most cases, the accuracy of English speech decreases as the SNR decreases. This is consistent with the conclusion that the mother tongue is the other language.