Method and device for recognizing natural speech

Abstract

The invention discloses a method and a device for recognizing a natural speech, and relates to a speech recognition technology, so as to solve the problem of low speech recognition success ratio because of a keyword mode. The method comprises the steps as follows: obtaining pinyin corresponding to a speech message input by a user; carrying out word segmentation treatment on the pinyin by a dictionary set in advance to obtain a segmented word pinyin string; searching a word to be recognized corresponding to the word pinyin string from the dictionary, and searching a target information database to obtain the target information which has the highest matching degree with the word to be recognized according to the word to be recognized, wherein the dictionary is used for storing a target word to be recognized by speech and the pinyin corresponding to the target word. The technical scheme provided by the embodiment of the invention can be applied to information service systems for navigation, song requesting, linkman inquiry and the like.

Description

technical field [0001] The invention relates to speech recognition technology, in particular to a natural speech recognition method and device. Background technique [0002] In the field of speech recognition, speech recognition technologies are different for different languages. For example, for English, the words in the sentence to be speech recognized are all composed of letters in the 26 alphabets. When performing speech recognition, The speech recognition system only needs to recognize the letters in the sentence to recognize the text information corresponding to the speech information. [0003] The biggest difference between Chinese and English is that the number of Chinese characters is relatively large. At present, the total number of Chinese characters has exceeded 80,000, of which about 3,500 are commonly used Chinese characters. Facing such a huge Chinese character library, traditional speech recognition technology is based on For keywords, the voice recognition ...

AI Technical Summary

Problems solved by technology

In order to improve the success rate and accuracy of speech recognition, it is necessary to store all expressions of the same thing in the vocabulary of the speech recognition system as much as possible, which makes the vocabulary of the speech recognition system very large in scale, inconvenient to maintain, and Due to the large vocabulary, the speed of speech recognition system is slow

In addition, because people's language expression forms vary greatly, with the development of the times, language expression methods are also constantly updated, and it is impossible to exhaust all the expression forms of the same thing in the vocabulary of the speech recognition system, making the use of keywords for speech recognition low success rate

[0005] Chinese patents with application numbers CN00130067.9, CN03123123.3, and CN03138149.9 also disclose technical solutions related to speech recognition. However, the above-mentioned technical solutions can only perform speech synthesis or convert speech into text, but cannot realize speech recognition. Speech is converted into text information recognition, and the above-mentioned technical solution is designed for English speech recognition. According to the above analysis, English language and Chinese language are quite different in terms of the number of words and grammatical characteristics, so that even if the above-mentioned technical solution is applied to Chinese Also can't carry out effective recognition in speech recognition, the success rate of speech recognition is lower; In the Chinese patent that application number is CN99813093.1, disclose a kind of interactive user interface that adopts speech recognition and natural language processing, although can realize The recognition of speech into text information, however, this technical solution is also designed for the English language. In the process of speech recognition, the influence of factors such as grammar needs to be considered, and it still cannot be effectively applied to Chinese speech recognition.

Images