让人与计算机自由地交谈,机器能听懂人讲话,是汉语语音识别技术最终将实现的目标。进入九十年代,语音识别方面的研究进一步升温。连续语音识别技术正趋于成熟。还出现了诸多实用化的研究方向。今后,将由连续语音识别发展到自然话语识别与理解,并着手解决语音识别中的一系列难题。难度虽然很大,但前景乐观。 Let people talk freely with the computer, the machine can understand people speech, is the Chinese speech recognition technology will eventually achieve the goal. Into the nineties, speech recognition research further warming. Continuous speech recognition technology is maturing. There are also many practical research directions. In the future, from continuous speech recognition to natural speech recognition and understanding, we will start to solve a series of problems in speech recognition. Although the difficulty is great, but the outlook is optimistic.