摘要 |
PURPOSE:To enable the determination of the capacity of a word already registered while providing the yardstick of the enunciation speed by displaying sound sections and cumulative time detecting the starting point of input voice through a comparator. CONSTITUTION:Input voice is rectified with a rectification smoothing circuit 4 via a noise removing LPF3 and compared with the threshold by a level comparator 7. When the threshold is exceeded, the comparison output is inverted to a high level and by detecting the input voice starting point and the sound section, the sound sections undergo a time counting with a sound section timer 11 through a CPU. Likewise, the time counting is performed for the soundless sections with a soundless section timer 12. In addition, displays 13 and 15 show accumulations of sound sections of words uttered immediately before and those thus far uttered respectively through drivers 14 and 16. The display of the sound sections gives a speaker the yardstick of the enunciation speed while enabling the learning of the capacity of words already registered into a recognition equipment to achieve an efficient utilization of the recognition equipment. |