Method based on voice translation text output

Abstract

The invention relates to a method based on voice translation text output. The method is characterized in that steps of original voice storage, original voice correction, phonetic transcription storage, voice text transformation, translated text storage and text storage of an original voice are carried out sequentially, so output is realized after the original voice is translated into a text signal. According to the method, not only can automatic correction on a dialect in an inputted original voice be carried out, but also due punctuations in the corresponding translation content can be automatically determined according to voice pause and rising and falling of intonation, till segment sentence modification is carried out to form a standard text, and thereby printing and output of the text or synchronous seeing and hearing of the text and the original voice are realized. The method is not only suitable for voice-text transformation communication of a common telephone set, a multimedia telephone set and an IP telephone, but also is suitable for terminal equipment, such as a radio, a recorder, a recording pen, a radio cassette player and a translator device which require voice-text transformation.

Description

technical field [0001] The invention belongs to the technical field of speech translation, in particular to a method for text output based on speech translation. Background technique [0002] With the continuous advancement of modern communication technology, digital technology, and software technology, communication equipment and communication terminals are becoming more and more diversified, personalized and high-end, and multifunctional communication equipment and communication terminals are convenient for users to operate and use. It is more and more favored by the public, especially in the application fields of modern private network communication, local area network communication and office automation. It has put forward high expectations for the development of terminal equipment with the function of real-time automatic translation of voice communication and text communication, but it is still basically is blank. [0003] Apple in the United States submitted a patent ...

AI Technical Summary

Problems solved by technology

However, these products still have obvious deficiencies: first, they can only send short message phrases, and the voice-to-text translation cannot be performed continuously due to restrictions such as the number of text messages; second, the accuracy of voice recognition is low, and dialects or non-standard pronunciation will cause errors. Conversion; third, punctuation marks and sentence breaks cannot be reflected; fourth, there is no text printout function, etc.

[0005] Chinese patent application 200710124172.X discloses "a method and mobile phone device for editing short messages by voice", which is controlled by a central processing unit and has a microphone that converts sound into voice signals, voice signals into text signals of short messages, and text The function of signal transmission to the display screen, but this method still has the following obvious shortcomings: first, it can only send simple sentences belonging to SMS phrases, and is not suitable for sending continuous long sentences; second, there is no sentence modification function, and continuous long sentences cannot be formed The third is that the application objects are limited to mobile phones, and the scope of application is very small

[0006] At present, in the use environment of ordinary telephones, multimedia telephones, IP telephone communications, etc., there are no products with commercial value that use the function of translating text from voice

Images