A method which suggests the extraction of features that are experimentally established as important to human perception of speech, e.g., voice onset times.; 一種(語(yǔ)音識(shí)別)方法,建議提取通過實(shí)驗(yàn)證明對(duì)理解人類語(yǔ)音而言是十分重要 的特性［征］,如聲音開始的時(shí)間,從而理解語(yǔ)音。參閱speech recognizer。