CN1805009A - Multi-language speech switching apparatus - Google Patents

Abstract

The invention relates to a multilingual voice exchange system in the field of communication technology. It is formed by a Chinese common speech input and output device, a Chinese electronic speech recognition system, a Chinese word stock, a French word stock, a French electronic speech recognition system, a Japanese word stock, a Japanese electronic speech recognition system, a Germany word stock, a Germany electronic speech recognition system, a Russian word stock, a Russian electronic speech recognition system, a language type transferring switch, a foreign language standard speech output and input device and a foreign language standard speech output and input transferring switch. It first uses the voice identifying system to transfer the inputted language into the local character and uses the translating software to translate the local character into the opposite character, and then it uses the language transferring software to transfer the opposite character into the sound and outputs in by the loudspeaker.

Description

A kind of multi-language speech switching apparatus

Technical field

The present invention relates to communication technical field, the particularly a kind of people's multi-language speech switching apparatus of open mutually that only can say Chinese or French, Japanese, German, Russian that makes respectively.

Background technology

Along with the development that deepens continuously of reform and opening-up, the raising day by day of Chinese international status, and after the China's entry into the WTO, the contacts of China and countries in the world are more and more many, and not only intergovernmental contacts are frequent, and the interchange in economy, the culture field is also more and more.Continuous opening along with tourist industry, there is the increasing foreigner to come China to visit, go sight-seeing, and along with the raising day by day of China's living standards of the people, also having increasing Chinese from now on goes on a tour abroad, and the biggest obstacle that language obstacle has become Chinese to exchange with the foreigner brings great inconvenience for work, life and trip.Therefore and will popularize foreign language in any one country nearly all is impossible, and competent country is all at the research machine translation mothod, but also all is on the computing machine intertranslation stage between two kinds of literal at present, reads voice by knocking keyboard.The biggest problem of mechanical translation is exactly speech recognition, and domestic and international existing speech recognition system, as the Via Voice that IBM releases, domestic sky letter speech recognition system, the logical speech recognition system of generation sound or the like, because vocabulary is little, can only the identification division entry maybe can only discern specific people's voice.For the voice of unspecified person, because the arithmetic capability of computing machine is limited, recognition performance is relatively poor.Therefore at present speech recognition technology mainly is used in voice control, phonetic dialing, high-grade speech-sound intelligent toy, intelligent sound telepilot, palm PC etc., does not also have to realize the device of the sound language intertranslation of computing machine between different language.

Summary of the invention

The object of the present invention is to provide a kind of energy can realize the sound language intertranslation of computing machine between languages such as Chinese and French, Japanese, German, Russian, and little, the portable multi-language speech switching apparatus of volume.

The objective of the invention is to be achieved through the following technical solutions: a kind of multi-language speech switching apparatus, it is articulated in computword intertranslation software and literary composition language switching software on the speech recognition system, realizes the sound language intertranslation of computing machine between Chinese and French, Japanese, German, Russian languages.At first will in computing machine, be converted into national literal through the standard language sound of microphone input by speech recognition system, by translation software national character translation is become the other side's literal, become sound language to emit the other side's text conversion by literary composition language switching software again by loudspeaker.It is made of Chinese mandarin input and output unit, Chinese electronic speech recognition systems, Chinese vocabulary bank, French dictionary, French electronic speech recognition systems, Japanese dictionary, Japanese electronic speech recognition systems, German dictionary, German electronic speech recognition systems, Russian dictionary, Russian electronic speech recognition systems, languages switch, the words output of foreign language standard and input media and foreign language standard words output input switch.With in to translate Russia be example, its workflow is: by microphone Chinese mandarin is imported → by Chinese electronic speech recognition systems voice are changed into Chinese character → go out to be fit to translate into the statement of Russian → be forwarded to Russian dictionary by the languages switch by the recognition and verification of Chinese vocabulary bank in computing machine, confirm Russian statement that an implication presses close to the most → convert Russian statement to voice → be forwarded to loudspeaker with the output of Russian standard words with the contrast of Russian dictionary by output input switch by Russian speech recognition system.Otherwise, workflow during Russia translates is: with Russian standard words input → be forwarded to Russian electronic speech recognition systems by output input switch voice to be changed into Russian statement → go out to be fit to translate into the statement of Chinese → be forwarded to Chinese vocabulary bank through the languages switch by the recognition and verification of Russian dictionary in computing machine, the Chinese statement that an implication presses close to the most → Chinese statement is converted to voice → export with Chinese mandarin by loudspeaker through speech recognition system is confirmed in contrast by microphone.In the device, Chinese and French, Japanese, German, Russian electronic speech recognition systems all adopt continuous electronics speech recognition technology.Chinese vocabulary bank, French dictionary, Japanese dictionary, German dictionary, Russian dictionary constitute by word library, phrase storehouse and short sentence commonly used storehouse.

The present invention is owing to adopting very big Chinese vocabulary bank of continuous speech recognition technology and vocabulary and French, Japanese, German, Russian dictionary, it is higher to know the sound recognition capability, only understand the people of Chinese or French, Japanese, German and Russian, can realize open by this device.And this device volume is little, is easy to carry.

Description of drawings

Fig. 1 is the multi-language speech switching apparatus functional-block diagram.

Embodiment

A kind of multi-language speech switching apparatus is made of Chinese mandarin input and output unit, Chinese electronic speech recognition systems, Chinese vocabulary bank, French dictionary, French electronic speech recognition systems, Japanese dictionary, Japanese electronic speech recognition systems, German dictionary, German electronic speech recognition systems, Russian dictionary, Russian electronic speech recognition systems, languages switch, the words output of foreign language standard and input media and foreign language standard words output input switch.With in to translate Russia be example, its workflow is: at first import Chinese mandarin with Chinese mandarin by input media (microphone), for example " how do you do? " these voice change into Chinese character by high capacity Chinese electronic speech recognition systems with voice, pass through again by Chinese word library, the Chinese vocabulary bank identification that phrase storehouse and short sentence commonly used storehouse constitute, on the screen of switch, demonstrate " how do you do? " " you are good? " " you howl? " etc. multiple corresponding statement, then with manual mode select be fit to translate into Russian and implication press close to the most Chinese words and phrases " how do you do? " confirm input, be forwarded to by Russian word library through the languages switch then, " 3APABCTBy is confirmed in Russian dictionary contrast that phrase storehouse and short sentence commonly used storehouse constituteTE, " Russian statement, and be presented on the screen of switch, pass through Russian electronic speech recognition systems at last again with Russian statement " 3APABCTBy

TE " convert voice to and be forwarded to loudspeaker with the output of Russian standard words by output input switch.The course of work during Russia translates is similar.

Claims (2)

1, a kind of multi-language speech switching apparatus, it is articulated in computword intertranslation software and literary composition language switching software on the speech recognition system, realize the sound language intertranslation of computing machine between different language, its technical characterictic is: it is by Chinese mandarin input and output unit, the Chinese electronic speech recognition systems, Chinese vocabulary bank, the French dictionary, the French electronic speech recognition systems, the Japanese dictionary, the Japanese electronic speech recognition systems, the German dictionary, the German electronic speech recognition systems, Russian dictionary, Russian electronic speech recognition systems, the languages switch, words output of foreign language standard and input media and foreign language standard words output input switch constitute, at first will in computing machine, be converted into national literal through the standard language sound of microphone input during interchange, and become sound language to emit the other side's text conversion by literary composition language switching software again by loudspeaker by speech recognition system.

2, a kind of multi-language speech switching apparatus according to claim 1, its technical characterictic is: Chinese, French, Japanese, German, Russian electronic speech recognition systems all adopt the continuous speech recognition technology, and Chinese, French, Japanese, German, Russian dictionary constitute by word library, phrase storehouse and short sentence commonly used storehouse.