摘要 |
<p>PURPOSE:To input an image recording reservation through voice by integrating a voice input part, a feature extracting part, a recognizing/deciding part, and a language processing part in a video tape recorder. CONSTITUTION:When a speaker speaks an item to be reserved while depressing a voice input key 8 provided on a video tape recorder, a voice signal converted into a digital signal is converted into a frequency dimension by the feature extracting part 2. A feature parameter extracted from a voice spectrum to be a converted signal is labelled by the 256-th order vector quantization in the recognizing/deciding part 3 and which phoneme is spoken as voice is recognized and decided. A phoneme string obtained by said method is sent to a language processing part 4 and converted into a meaning language such as words and clauses and the meaning is analyzed and converted into a meaning code, which is sent to a central control part 5. Consequently, the image recording of a required program can be reserved by voice.</p> |