JPS592096A - Language converter - Google Patents

Abstract

(57) [Summary] This bulletin contains application data before electronic filing, so abstract data is not recorded.

Description

DETAILED DESCRIPTION OF THE INVENTION (1) Technical Field of the Invention The present invention is directed to a language conversion device that automatically performs language conversion.

(2) Background of the technology As a language conversion system, for example, machine translation, which converts English words into Japanese by interpreting each word and syntax, is a literal conversion and there are restrictions depending on the target of conversion, but it exists. . There are also systems that convert language into speech.

However, recently there has been a demand for so-called automatic interpreters that convert audio input in one language into audio output in another language.

(3) Object of the Invention An object of the present invention is to provide a language conversion device as an automatic interpreter that converts audio input in one language into audio output in another language.

(4) Structure of the Invention In order to achieve the above-mentioned objects, the present invention provides: a language recognition means for syntactically recognizing a speech signal of a language; and converting the syntactically recognized language into another language by word and syntactic interpretation. There is provided a language conversion device characterized by comprising a language conversion means for converting the language, and a voice conversion means for converting the converted language into an audio signal.

(5) Examples of the invention Hereinafter, the present invention will be explained in detail with reference to the drawings.

FIG. 1 is a block circuit diagram showing an embodiment of a language conversion device according to the present invention. In FIG. 1, 1 is a central processing module (processor) that controls the entire system, and 2 is a demodulator that demodulates, for example, an English speech transmission signal SIN sent from a transmission path into a speech signal. . The speech signal from the demodulator 2 is syntactically recognized by a language recognition module 3.

Next, the language conversion module 4 interprets the words and syntax and converts them into, for example, Japanese. This is then converted into an audio signal by the audio conversion module 5 and further voiced by the audio generator 6 into the audio S. U
, becomes. Note that 31°41 to 44.51.52 are dictionaries (tables) necessary for the operation of each module 3 and 4°5, and their contents are usually stored on a large-capacity and high-speed storage medium (
(not shown) and is controlled by the central processing module 1. Still, each module 3, 4.5 is implemented as a ROM (LSI) in module units and is controlled by the central processing module 1.

The operation of the circuit shown in FIG. 1 will be explained with reference to the flowchart shown in FIG. The flowchart in FIG. 2 starts at step 201 and ends at step 213. Steps 202 to 204 indicated by ■ indicate the operation of the language recognition module 3.
Ste indicated by ■.

Steps 205 to 210 indicate the operations of the language conversion module 4, and steps 211 and 212 indicated by ■ indicate the operations of the voice conversion module 5.

That is, in step 202, the language recognition module 3 converts the audio signal to, for example, 6.1.4SLIIJLI! uuuuu9uPue-
Separate and recognize each syllable like +N. Next, in step 203, 1sqj? '. Each word is reconstructed as buPeN, and in step 204, each word is compared with the word dictionary $31 and recognized as a word syntax such as Th1suisL, IaUpen.

Next, the language conversion module 4 performs preprocessing to detect keywords in step 205.

For example, in the preamble, the keyword pen is detected. Next, in step 206, the keyword is referred to the idiom dictionary 41 to recognize syntax, idioms, etc. In step 207, each word is recognized for each part of speech.

That is, with reference to each part-of-speech dictionary 42, each word is classified as conjunction 1.
Nouns 2 Recognize as pronouns, prepositions, verbs, adjectives, and other parts of speech.

Next, in step 208, the syntax no! turn dictionary 4
3. Perform syntactic pattern recognition with reference to 3. For example, recognize the above sentence as a syntactic pattern of pronoun ten-do-a + preposition ten-noun. In step 209,
Word conversion processing is performed for each word with reference to the language correspondence dictionary 1t44. For example, convert to . Next, in step 210, again the syntax t'? Syntax processing is performed using the turn dictionary 43.

That is, This is one in of J1. and process it.

Next, the voice conversion module 5 performs pitch adjustment processing of the syntax in Stella 7' 211 with reference to the pitch adjustment pattern dictionary 51. In other words, it is processed as follows. Next, in step 212, the speech dictionary 52 is referred to to perform word speech generation processing, that is, /-/ ゝ\ / ゝ\Korewa
LJ 1ppon u noupenUdlls
u and sends it out as an audio signal, which audio signal is to be converted into audio by the audio generator 6. (4) As described in detail, according to the invention, - the audio of a language is Can convert audio to other languages. In other words, you can listen to a foreign language as your own language without an interpreter.

[Brief explanation of drawings]

FIG. 1 is a circuit diagram showing an embodiment of a language conversion device according to the present invention, and FIG. 2 is a flowchart for explaining the operation of the circuit shown in FIG. 1: Central processing module 2: Demodulator 3: Language recognition module 4: Language conversion module 5: Speech conversion module 6: Speech generator Patent applicant Fujitsu Limited Patent agent Akira Aoki Patent attorney Nishidate Oro Patent Attorney 1) Yukio Patent Attorney Akiyuki Yamaguchi's t'i Bow IF (Portrait;!, -,'Yashi na L) Procedural Amendment (Voluntary) July 79, 1982 Kazuo Wakasugi, Commissioner of the Japan Patent Office1, Indication of the case 1982 Patent Application No. 1106302, Name of the invention Language conversion device3, Person making the amendment Relationship with the case Patent applicant name (522) Katsushitsu Stock Company 4, agent (3 others) 5. Drawings to be amended (Fig. 1, Fig. 2) 6. Contents of amendment Official drawings will be left behind. (However, the contents cannot be changed.) 7. List of attached documents

Claims (1)

[Claims] 1.-Language recognition means for syntactically recognizing a speech signal of a language;
A language conversion device comprising: a language conversion means for converting the syntactically recognized language into another language by word and syntactic interpretation; and a speech conversion means for converting the converted language into an audio signal.