With the widespread growth in the use of digital computers, there has been an increasing need for a man to be able to communicate with machines in a manner more naturally suited to humans.
The realization of this need has motivated a great deal of research in automatic speech recognition by computer.
This thesis describes a speaker-independent recognition system for separately uttered Korean digits.
Parameters such as ZCR (Zero-Crossing-Rate) and run-length of signal/derivative-signal are normalized to be independent of the fundamental frequency of each speaker's own, using the estimate of fundamental pitch-period for the voice-signal to be analyzed.
Digits are discriminated into voiced-frames and/or unvoiced-frames at the beginning portion of the voice signal, using ZCR and intensity measurements.
Then, voiced-frames are classified further using the normalized parameters of ZCR and run-length.
Recognition experiment for this system is carried out for 50 digits-utterences by one female speaker and the recognition ratio is obtained to be 98%.