| 摘要 |
<p>PURPOSE:To automatically and quantitatively perform voice recognizing inspections by performing a voice recognizing process on one of registered voice patterns which are registered by patterning a prescribed number of voice signals for tests by means of a voice recognition processing means after again outputting the pattern in voice signals. CONSTITUTION:The recognizing code number corresponding to one of the voice patterns registered in a voice recognition processing means 4 is outputted as a voice signal by means of a voice synthesization processing means 3. The means 4 compares the voice pattern with the voice patterns registered in the means 4 for similarity and discriminates whether or not the corresponding code number coincides with a selected recognizing code by comparison after selecting the corresponding recognizing code number. The propriety of a voice recognizing section can be automatically and quantitatively discriminated without inputting the voice of an inspector by connecting the inputting section of the means 4 with a voice inputting means 6 and outputting section of the voice synthesization processing means to a voice outputting means 7.</p> |
