| 摘要 |
A baseband modem and method for voice recognition and a mobile communication terminal using the baseband modem and method are disclosed. A speech recognition rate may be increased by selecting a sampling rate suitable for speech recognition and portions of the speech recognition process may be implemented in hardware. The present invention includes an audio codec modulating a received voice signal using either a sampling rate for speech recognition or a sampling rate for voice communication. A feature vector extraction block extracts one or more feature vectors from the modulated voice signal and a speech recognition block performs speech recognition using an extracted feature vector when the voice signal is determined as a voice command. A vocoder vocodes an output of the audio codec when the voice signal is determined as voice communication. |
