开发并评估一个中文计算机语音测试系统。

Capital Medical University, PR China.

Comput Biol Med. 2011 Mar;41(3):131-8. doi: 10.1016/j.compbiomed.2011.01.002. Epub 2011 Jan 31.

OBJECTIVES

This study reports the development and evaluation of a Computerized Mandarin Speech Test System (CMSTS).

METHODS

Taking into account the rules for developing speech materials and the unique linguistic characteristics of Mandarin, we designed and digitally recorded a set of materials comprised of seven lists of monosyllabic words, nine lists of disyllabic words, and fifteen lists of sentences with a high degree of subject familiarity. The CMSTS was developed with Visual Studio 2008, Access 2003 and DirectX 9. The system included five functions: listener management, a speech test, list management, data management, and system settings. We used the system to measure the speech recognition threshold (SRT) of 76 participants with normal hearing (age range: 20-28 years), and measured performance-intensity functions (PI) for all stimuli.

RESULTS

The SRT results were in accord with pure-tone results obtained by pure-tone audiometry. In a speech recognition score (SRS) test, changing the presentation level had the strongest effect on sentence recognition, followed by the presence of disyllabic words. Monosyllabic words were least affected by changes in presentation level. The slopes of the linear portion of the PI using the system were in accord with the findings of previous studies using audiometers and CDs with similar materials.

CONCLUSION

The CMSTS has sufficient sensitivity, and can facilitate the wider use of speech audiometry in Chinese audiology clinics.

目的

本研究报告了计算机普通话语音测试系统（CMSTS）的开发和评估。

方法

考虑到语音材料开发规则和普通话独特的语言特点，我们设计并以数字方式录制了一套材料，包括七个单音节词列表、九个双音节词列表和十五个具有高度主题熟悉度的句子列表。CMSTS 是使用 Visual Studio 2008、Access 2003 和 DirectX 9 开发的。该系统包括五个功能：听众管理、语音测试、列表管理、数据管理和系统设置。我们使用该系统测量了 76 名正常听力参与者（年龄范围：20-28 岁）的语音识别阈值（SRT），并测量了所有刺激的性能-强度函数（PI）。