ON-LINE AUTOMATIC RECOGNITION OF SPOKEN SOUNDS BY A PATTERN MATCHING METHOD
用图样匹配法在计算机上自动识别语音

A model for speech recognition by computer is suggested. In which a new method of pattern normalization in the time domain is presented. A program for recognizing spoken sounds has been developed on a Varian 620/L computer. Finally, a set of ex-perimental results is given, the correct recognition rate being grater then 99%. It is shown that this new method can be used in certain automatic control systems.