| 术语 | speakerdependent /independent systems |
| 释义 | speakerdependent /independent systems 说话者有关/无关型系统 There are many ways of dividing voice systems. One way is deciding if they must be “trained” by individual users. All speech systems are either speaker dependent or speaker independent. A speaker dependent system recognizes only those voice it has been trained to recognize; it works best in applications with only a few users. To train a system, each speaker must repeat every active vocabulary word or phrase several times to store a recognizable digitalized voice pattern in memory. When that individual subsequently uses the system, each spoken word is compared with the stored patterns. These systems are unaffected by voice inflections, regional accents, and other speech changes, are easy to learn, and can be used over telephone lines because they are designed to accommodate both noise on the line and background noise. Speakerindependent systems make larger vocabularies available to fewer users, typically, ten or less. They are sensitive to minor voice changes, must be trained by each successive us er,and can be affected by noisy telephone lines or by background noise. 有多种区分语音系统的方法。一种区分方法就是根据这些系统是否必须由个别用户进行“训练”来定。 所有语音系统或是说话者有关型或说话者无关型的。说话者有关型系统只能识别那些对其进行 过训练的说话者,它可以很好地用于只有少数几个用户的应用中。要训练一个系统,每个说话 者必须对每个常用词汇或短语重复说几次,以便在内存中存储一个可识别的数字化的语音 模式,当某个人以后使用该系统时,说的每一个词都要同存储的语音模式进行比较。 这种系统不受音词、地方口音或其它语音变化的影响,且易于学习,并可通过电话线路使用, 而不受线路噪声和外来噪声的影响。说话者无关型系统可使较多的词汇适用于少数几个用户, 一般在十个之内,这种系统对声音的变化很敏感,必须由逐个用户进行训练,且易受电话线路 和外来噪声的影响。 |
| 随便看 |
|
计算机英汉双解词典包含21137条计算机术语英汉翻译词条，基本涵盖了全部常用计算机术语的翻译及用法，是计算机学习及翻译工作的有利工具。