KR20050052943A - Key-word based speech translation system for travelers and the method - Google Patents

Abstract

The present invention communicates with travelers by uttering a minimum keyword using a mobile communication terminal in a face-to-face situation where the other party does not know each other's language, and then confirms the native language character recognition result displayed on the screen and the translated characters in the other country's language. It relates to a keyword-based speech translation system and a method for smoothing the.

The method of the present invention comprises the steps of: setting a first language and a second language, which are languages of two speakers, receiving a first language keyword voice from a user, and transmitting the first language keyword voice to a voice translation server; Converting the first language character into a first language character and transmitting the same to the first language character, and confirming whether or not the voice recognition result displayed on the mobile terminal as the first language character is correct, Translating the language text into the second language text, and the mobile terminal receives the translation result and outputs the second text.

Description

Keyword-based Speech Translation System for Travelers and its Method {Key-word based Speech Translation System for Travelers and the Method}

The present invention relates to a voice translation system and method for a traveler, and more particularly, in a face-to-face situation in which a traveler does not know each other's languages in the course of traveling abroad, a mobile phone, a PDA, or the like. A keyword-based speech translation system for travelers that facilitates the communication of travelers by making a minimum number of keywords using a mobile communication terminal and then confirming the results of the national character recognition on the screen and the translated characters in the language of the other country. And to a method thereof.

In the case of an automatic interpretation system, speech recognition and speech synthesis techniques are generally used. Recently, an attempt has been made to define an intermediate language for mediating different languages and to provide automatic interpretation. The automatic interpretation system undergoes processes such as speech recognition, sentence interpretation and translation into intermediate languages, generation of target language and speech synthesis.

However, since the existing system has to construct and recognize a large vocabulary, the speech recognition rate is not satisfactory, and the result of interpretation based on this is also unreliable.

In addition, if the recognition error occurs due to the nature of the automatic interpretation in pursuit of real-time can not determine whether the error can be reproducible and corrected. That is, the conventional automatic interpretation system is inefficient to use for communication in the face-to-face situation, and there is a problem that causes unnecessary waste of communication costs.

Meanwhile, a method of performing domain recognition, sentence recognition, and outputting a target sentence after embedding a speech target sentence in a portable terminal device may be attempted, but this method also cannot escape the limitation of speech recognition error, and communicate with each other. For this problem, the communication process becomes rather complicated, such as reselecting a desired sentence among many candidate sentences.

In addition, as a way to avoid the risk of voice recognition errors and to support communication between Koreans and foreigners, a separate interpreter has been proposed at the call center to provide direct service in the middle of foreigners, but this procedure is cumbersome. In addition, there are many human and material costs associated with the use of interpreters.

Accordingly, the present invention is to solve the above-mentioned problems, the object of the present invention is to enable communication with a minimum of keywords (Key-word) to significantly reduce the number of vocabulary of speech recognition targets as a result It is to provide a tourist keyword-based speech translation system and method for overcoming ambiguity in communication as well as reducing errors in speech recognition due to large-scale vocabulary search.

Another object of the present invention is to reduce the time between speech and speech by displaying the voice translation results in letters, and to reduce the speech recognition and translation target vocabulary by inserting a region selection function, as well as to eliminate vocabulary such as proper nouns. The present invention provides a keyword-based speech translation system and method for travelers that can enhance communication by excluding and utilizing intelligence and situational knowledge among users.

The keyword-based voice translation system for travelers for achieving the object of the present invention, a language setting module for setting the first language input by voice and the second language to be translated and displayed by the text, and the user's first language keyword voice After inputting and extracting the feature information through the voice recognition pre-processing process, the voice input module for transmitting the result data to the voice translation server, the voice recognition result of the voice translation server is displayed in a first language character, The translation result is a mobile terminal including a character display module for outputting and displaying the second language characters; And a speech recognition module for recognizing user speech data from the mobile terminal and converting the user speech data into a first language character, a translation module for translating the first language character of the recognition result into a second language character, and the first speech recognition result. And a voice translation server including a text sending module configured to send one language text and second language text data of the translation result to the mobile terminal.

In addition, a keyword-based speech translation method for travelers to achieve the object of the present invention, comprising: setting a first language input by voice and a second language to be translated and displayed by text; Receiving a first language keyword voice of a user through the mobile terminal and transmitting the voice data to the voice translation server; Recognizing the voice data of the user by the voice translation server, converting the voice data into a first language character, and transmitting the recognition result data to the mobile terminal; Outputting and displaying the voice recognition result as a first language character on a mobile terminal, and confirming whether or not the voice recognition is correct by a user; And the voice translation server translating the first language character of the recognition result into a second language character, and the mobile terminal receiving the translation result data and outputting the second language character to output the second language character.

In general, conversations in face-to-face situations can utilize paralinguistic features such as gestures and facial expressions, allowing communication between users of different languages with a small vocabulary. In other words, if the meanings of the core vocabulary of the conversation such as nouns or noun phrases, verbs or verbs, and instructional adverbs and interrogative adverbs can be communicated with each other, communication can be easily achieved. In the present invention, the key-word refers to such a core vocabulary.

According to the present invention, since a key-word can be formed only of some nouns, verbs, and indication adverbs except proper proper names among the existing tourist corpus, speech recognition errors due to large-scale lexical search can be drastically reduced.

Hereinafter, exemplary embodiments of the present invention will be described in detail with reference to the accompanying drawings.

1 schematically illustrates a block configuration and processing procedure of a voice translation system based on keywords according to the present invention.

The tourist key-word speech translation system of the present invention recognizes the source language (eg, A language) as a character of the source language by using the mobile terminal 100 and displays it to the user, and the target language ( For example, as a system for translating into a letter of B language), as shown in FIG. 1, the mobile terminal 100 is largely comprised of a communication network and a voice translation server 200.

Here, the mobile terminal 100 is a terminal having a wireless data transmission and reception function, can be implemented as a dedicated terminal for the voice translation service of the present invention, and also functions for the present invention in a mobile communication terminal such as a mobile phone or PDA Modules may be implemented in the form of hardware or software addition.

In the present invention, the mobile terminal 100 receives the voice of the original language from the user and transmits it to the voice translation server 200 through a communication network, and receives the recognition result from the server to the text of the source language Display the output. Further, translation data of the recognition result is received from the server, and output is displayed in characters of the target language (B language). In addition, the mobile terminal 100 to display directly by using a key button in the case of numbers or special characters.

 To this end, the mobile communication terminal includes a language setting module, a region selection module, a voice input and voice recognition preprocessing module, a key input module, a positive / false check module, a character display module, and the like in hardware or software.

Here, the language setting module receives the language of both the dialogues from the user. That is, the user can designate which language is the source language and the target language. In addition, the user can switch between the source language and the target language through simple key manipulation of the user. For example, if the source language is set to 'Korean' and the target language is set to 'English', the Korean user's key-word voice is displayed in English characters. By translating the source language and the target language, the English voice of the other party can be translated into Korean characters.

The area selection module is a module for selecting a situation where a current conversation is being made by a user, and selecting a conversation place such as a restaurant, an airport, a shop, a tourist attraction, or a conversation content such as a transaction, a tour guide, and a lodging. Such an area selection module may manually select a conversation area by a user or perform an automatic area selection function.

In addition, the voice input and voice recognition preprocessing module receives a user's native language voice through a microphone means, and extracts the features of the voice data to perform a preprocessing process for voice recognition, and the resulting data to the voice translation server ( 200). In this case, the speech recognition preprocessing may be performed by the voice translation server, not the mobile terminal.

The key input module receives a number or a special character for displaying an amount or a distance by the user's key button selection and outputs the result directly on the terminal without a separate translation process.

In addition, the correct / false confirmation module is a module for confirming whether the voice recognition result is correct for the primitive language from the user, the character display module to display the text data according to the voice recognition result and the translation result in the character of the language Module.

Meanwhile, the voice translation server 200 transmits a network interface module 210, a voice recognition module 220, a translation module 230, and text transmission to perform an interface function with a user's mobile terminal 100 through a communication network. Module 240 and the like.

The voice recognition module 220 recognizes a user's voice through language setting information received from the mobile terminal 100 and feature information of the user's key-word voice, and outputs the recognition result as a source language character. In the case of a homonym, the recognition results are displayed to the user so that the user can select the correct recognition character.

The translation module 230 translates the source language character that is the recognition result into the character of the target language designated by the user. The present invention targets words or phrases that are key elements in conversational content, such as nouns or noun phrases, verbs or verbs, and instructional adverbs and interrogative adverbs, so that accurate translation can be easily performed even between languages having different language structures. have.

In addition, the text transmission module 240 transmits the source language text that is the voice recognition result or the target language text that is the translation result to the user mobile terminal 100.

Next, a brief description will be made of a voice translation processing procedure according to the present invention with reference to FIG.

First, the user sets up a language and selects a region, and then speaks key words necessary for dialogue in his or her native language. Accordingly, the mobile terminal 100 performs a preprocessing process for speech recognition such as extracting feature information of the user's voice signal, and then transmits the data to the voice translation server 200 together with the language setting and region selection information. do.

The voice translation server 200 checks the language setting and the conversation area information, recognizes the user's voice as a character of the original language through the voice recognition module 220, and then transfers the recognition result data to the mobile terminal 100. Send it out. Accordingly, the mobile terminal 100 displays the recognition result to the user in original language characters, and the user transmits an error of the displayed recognition result to the server. At this time, if there is an error in the recognition result, the user goes through a recurrence process.

If the meaning of the speech recognition is very similar (eg, inconsistency between the terminology and the tense), it is advisable to carry out the recurrence process only when the meaning is completely different.

Meanwhile, the voice translation server 200 translates the recognized source language Key-Word character into the character of the target language set by the user and transmits the translation result data to the mobile terminal 100. The translated text of the target language is displayed on the mobile terminal, and the counterpart can check the meaning of the user key-word voice by looking at it.

2 is a flowchart illustrating a keyword-based voice translation process for travelers according to the present invention.

Referring to FIG. 2, first, a source language and a target language are set on a terminal, and a conversation area according to a current conversation situation and a place is selected.

At this time, the conversation area is classified into a situation or a place that is frequently generated by a traveler such as an airport, a restaurant, a store, a tourist destination, a lodging, and the like.

Then, an input mode for text or voice is selected (S202).

In this case, when the user wants to display a number or a special character, the user selects a character input mode and directly displays the number or a special character on the terminal by selecting a number or a special character for an amount or distance through key selection (S209).

On the other hand, if translation of the key word speech is required, the user selects a voice input mode and then inputs a key-word voice in his or her native language, and the mobile terminal 100 performs a voice recognition preprocessing process. After that, the result is transmitted to the voice translation server 200. (S203)

Also, the voice translation server 200 recognizes the voice data received through the voice recognition module 220 and converts the voice data into text of a source language set by the user (S204).

Then, the recognition result data is sent to the mobile terminal 100 of the user, and the mobile terminal 100 outputs and displays it in the corresponding original language character (S205).

The user checks the displayed key word character to see whether the voice spoken by him or her is correctly recognized, and if the key word of the user is recognized in a completely different meaning, the user performs a recurrence process (S206).

On the other hand, if there is no error in the recognition result, the voice translation server 200 translates the result data recognized as the source language character through the translation module 230 to the target language character, and the translation result data Is transmitted to the mobile terminal 100. (S207)

Accordingly, the mobile terminal 100 outputs the translation result as a character of the target language (ie, the counterpart language), and the counterpart recognizes the user's intention by looking at it (S208).

In addition, if the voice input of the other party is necessary for mutual exchange of doctors, the source language and the target language are switched through simple key manipulation (eg, by using a hot key), and then the processes of steps S202 to S208 are performed. Word displays the character of the user language for the voice input output (S210).

As described above, the keyword-based speech translation system for travelers and the method according to the present invention can reduce the cost of using the communication network and delay the turn between conversations because the translation results are shown in text without undergoing speech synthesis. Significantly reduce your time for natural communication.

In addition, the recognition result of the speaker's spoken word can be confirmed through a text message, so that translation errors due to recognition errors can be prevented in advance, and key-words, not sentences, are translated and recognized. There is a significant reduction in vocabulary.

What has been described above is just one embodiment for implementing the keyword-based speech translation system and method for travelers according to the present invention, the present invention is not limited to the above-described embodiment, it is claimed in the claims Without departing from the gist of the present invention, any person having ordinary knowledge in the field of the present invention will have the technical spirit of the present invention to the extent that various modifications can be made.

1 is a block diagram of a keyword-based speech translation system according to the present invention.

2 is a flowchart illustrating a keyword-based speech translation process according to the present invention.

<Description of the symbols for the main parts of the drawings>

100: mobile terminal 200: voice translation server

220: speech recognition module 230: translation module

240: character transmission module

Claims (8)

A mobile terminal with text display function for multiple languages and a voice translation server with voice recognition and text translation functions for multiple languages are interconnected through a communication network, providing keyword-based translation of voices between speakers in the language of the other language. In the voice translation method, (a) setting a first language input by voice and a second language to be translated and displayed by text; (b) receiving a first language keyword voice of a user through the mobile terminal and transmitting the voice data to the voice translation server; (c) recognizing the voice data of the user in the voice translation server, converting the voice data into a first language character, and transmitting the recognition result data to the mobile terminal; (d) outputting and displaying the voice recognition result in a first language character on a mobile terminal, and confirming whether the voice recognition is correct or not; And (e) the voice translation server translating the first language character of the recognition result into a second language character, and the mobile terminal receiving the translation result data and outputting the second language character to output the second language character. Keyword-based voice translation for travelers. According to claim 1, wherein step (b), A keyword-based speech translation method for travelers, characterized in that the user receives only the key words or phrases necessary for the conversation as the first language voice of the user. The method of claim 1, wherein step (d) In the case of a homonym having a plurality of recognition candidates for one voice, the recognition candidates are outputted to the user and one of the candidates is selected by the user. The method of claim 1, wherein the keyword-based voice translation method for travelers is provided. In the case of numbers and special characters, the keyword-based speech translation method for travelers, characterized in that the input directly displayed through the operation of the key button on the mobile terminal without access to the voice translation server. The method of claim 1, wherein the keyword-based voice translation method for travelers is provided. Prior to inputting the first language voice of the user, a conversation area for a place or a situation where a conversation is currently being made is set, and the voice translation server gives priority to a meaning related to the conversation area in recognizing the user's voice. A keyword-based voice translation method for travelers. In a voice translation system in which a mobile terminal and a voice translation server are interconnected through a communication network to translate a conversational voice between users into characters of a counterpart language, The mobile terminal, A language setting module for setting a first language input by voice and a second language to be translated and displayed by text; A voice input module for receiving a first language keyword voice of a user, extracting feature information, performing a voice recognition preprocessing process, and transmitting the result data to the voice translation server; The voice recognition result of the voice translation server is displayed in a first language character, the translation result includes a character display module for outputting display in a second language character, The voice translation server, A voice recognition module for recognizing user voice data from the mobile terminal and converting the user voice data into a first language character; A translation module for translating the first language character of the recognition result into a second language character; And a text sending module for sending the first language text of the speech recognition result and the second language text data of the translation result to the mobile terminal. The method of claim 6, wherein the mobile terminal, It further includes a region selection module for setting the dialogue area for the place or situation where the current conversation is being made, The voice recognition module of the voice translation server, When the voice data of the user is recognized in a plurality of meanings, the keyword-based speech translation system for travelers, characterized in that the user voice is given priority to the meanings related to the conversation area. The method of claim 6, wherein the mobile terminal, In the case of numbers and special characters, the keyword-based speech translation system for travelers, characterized in that the input directly through the key button operation on the mobile terminal without being connected to the voice translation server.