语音识别技术将人工智能赋予计算机,使它能够听懂人类语言,并迅速把起声音变成文字,真正实现人际交流。语音产品发展到今天,衡量标准日趋一致:连续语音还是非连续语音连续音才自然,输入速度也比较快;是特定语音还是非特定语音特定语音往往要长时间训练,也无法在公共场合使用,非特定语音技术不管男女老幼的声音都可以识别,即便是有轻微的口音也无妨;词条量的大小虽然日常生活中常用的词汇最多不超过一两万个,但每个专业的术语不少于几万条; Speech recognition technology gives artificial intelligence to the computer, enabling it to understand human language and quickly turn it into words to truly communicate. Today’s voice-based products are becoming more and more unified in their measurement standards: continuous or discontinuous voice continuous sounds are naturally faster in input speed; voice-only or non-specific voice-specific voice tend to be trained for a long time and can not be used in public places, Non-specific voice technology, regardless of the voices of men, women and children can be identified, even if there is a slight accent it anyway; the amount of entries Although the daily life of the vocabulary used up to no more than one or two million, but each professional terminology Less than tens of thousands