WO2019214299A1

WO2019214299A1 - Automatic translation apparatus and method, and computer device

Info

Publication number: WO2019214299A1
Application number: PCT/CN2019/073534
Authority: WO
Inventors: 张立新
Original assignee: 深圳市沃特沃德股份有限公司
Priority date: 2018-05-08
Filing date: 2019-01-28
Publication date: 2019-11-14
Also published as: CN108899018A

Abstract

Disclosed by the present invention are an automatic translation apparatus and method, and a computer device, the method comprising: respectively acquiring first voice information and second voice information collected by a first pickup and a second pickup; comparing the magnitudes of the voice amplitudes of the first voice information and the second voice information; outputting the voice information having the greater voice amplitude as translation information. By means of automatically translating according to the magnitudes of voice information received by two sound pickups, steps of manual configuration are reduced, thereby improving user experience.

Description

Automatic translation device, method and computer device

Technical field

The present invention relates to the field of translation machines, and in particular, to an automatic translation apparatus, method, and computer apparatus.

Background technique

With the development of international trade and the global village, cross-border or cross-language communication is becoming more frequent. When the two sides of the communication do not understand each other's language, they often need to use the translation machine to communicate, but now on the market. The intelligent translation machine needs to set the translation language by pressing the button first, and since the speech recognition is more difficult to recognize the mixed language, it is necessary to press and hold the translation key of each side to speak and start the corresponding language translation function, so when the two people talk, two are needed. People take turns to hold down the translation keys of their respective sides. The translation keys may be on the main unit or on the headphones or other wearable devices. In the communication, the button operations need to be performed continuously. This operation is very unhumanized and the user experience is more user-friendly. difference.

technical problem

The technical problem to be solved by the present invention is to provide an automatic translation apparatus, method and computer apparatus in view of the deficiencies in the above background art.

Technical solution

The technical means adopted by the present invention to solve the technical problem is to provide an automatic translation apparatus, including:

a pickup assembly comprising a first mounted pickup and a second pickup, the first pickup for collecting the first voice information, and the second pickup for collecting the second voice information;

a controller component, coupled to the first pickup and the second pickup, for receiving and comparing the magnitude of the voice amplitude of the first voice information and the second voice information, and controlling the sound pickup component to output a voice information with a large voice amplitude;

The main processor component is connected to the pickup component for receiving one voice information with a large amplitude of voice output by the pickup component, and according to the predetermined first language and the second language, the voice information having a larger voice amplitude The translation is performed to generate translation information, and the first language and the second language respectively correspond to the first voice information and the second voice information.

In another aspect, the present invention also provides an automatic translation method for use in an automatic translation apparatus as described above, comprising:

Acquiring the first voice information collected by the first pickup and the second voice information collected by the second pickup;

Determining whether the voice amplitude of the first voice information is greater than the voice amplitude of the second voice information;

If yes, the first voice information is translated from the first language to the second language to generate the first translation information, otherwise the second voice information is translated from the second language into the first language to generate the second translation information, where the first language and The second language corresponds to the first voice information and the second voice information, respectively.

The present invention also provides a computer device comprising a memory, a processor, and a computer program stored on the memory and operable on the processor, the processor implementing the method of any of the above.

Beneficial effect

With the above technical solution, the present invention has at least the following beneficial effects: the embodiment of the present invention is provided with a first pickup and a second pickup that are reversely mounted, thereby respectively acquiring the first voice information and the second voice information, when the first user is facing When the first pickup is speaking and the second user is speaking to the second pickup, the voice amplitude of the first user's voice signal collected by the first pickup is greater than the voice amplitude of the first user's voice signal collected by the second pickup. Similarly, the voice amplitude of the second user's voice signal collected by the second pickup may be greater than the voice amplitude of the second user's voice signal collected by the first pickup, and the controller component determines the first voice information and the second voice. The magnitude of the speech amplitude of the information to determine whether the first user is speaking or the second user; if the first user is speaking, the controller component controls the pickup component to transmit the first voice information to the main processor component, the first The voice information is translated from the first language into the second language to generate translation information; if the second user is speaking, the controller component The pickup component transmits the second voice information to the main processor component, and the second voice information is translated into the first language by the second language to generate translation information, wherein the first language and the second language respectively correspond to the first voice information and the first voice information The second voice information is automatically translated according to the size of the sound received by the two-way pickup, which reduces the manual setting steps and improves the user experience.

DRAWINGS

1 is a schematic structural view of an embodiment of an automatic translation apparatus of the present invention;

2 is a schematic structural view of another embodiment of an automatic translation apparatus of the present invention;

3 is a schematic diagram showing the circuit structure of an embodiment of an automatic translation apparatus of the present invention;

4 is a schematic diagram showing the circuit structure of another embodiment of the automatic translation apparatus of the present invention;

Figure 5 is a block diagram showing the flow of an embodiment of the automatic translation method of the present invention;

FIG. 6 is a schematic block diagram of a module of an embodiment of a computer device according to the present invention.

The implementation, functional features, and advantages of the present invention will be further described in conjunction with the embodiments.

BEST MODE FOR CARRYING OUT THE INVENTION

Referring to FIG. 1 to FIG. 4, the present invention provides a technical solution: an automatic translation apparatus, including:

The pickup assembly 1 includes a first pickup 11 and a second pickup 12 which are oppositely mounted, the first pickup 11 is for collecting the first voice information, and the second pickup 12 is for collecting the second voice information;

The controller component 2 is connected to the first pickup 11 and the second pickup 12 for receiving and comparing the magnitude of the speech amplitude of the first speech information and the second speech information, and controlling the pickup component 1 to output a path with a larger amplitude of the speech. voice message;

The main processor component 3 is connected to the pickup component 1 for receiving one voice information with a large amplitude of the voice output by the pickup component 1, and has a large amplitude of the voice according to the predetermined first language and the second language. The voice information is translated all the way to generate translation information, and the first language and the second language respectively correspond to the first voice information and the second voice information.

In one embodiment, taking the user as the first user and the second user as an example, the first language and the second language respectively correspond to the languages of the first user and the second user, and before the conversation, the first user and the second user The language can be input separately. Of course, before the communication is performed, the first user and the second user can also speak the language wake-up words corresponding to the language of the first pickup 11 and the second pickup 12 respectively, and the present invention automatically translates The device may respectively identify the first language corresponding to the first pickup 11 and the second language corresponding to the second pickup 12 through a preset offline voice library, thereby predetermining the first language and the second language. Specifically, in a case where the first language of the first user is Chinese and the second language of the second user is English, the first pickup 11 and the second pickup 12 are reversely mounted to each other such that the first pickup 11 is in different directions. Unlike the voice amplitude of the sound signal collected by the second pickup 12, the first user is facing the first pickup 11 and the second user is facing the second pickup 12.

When the first user speaks, the first pickup 11 and the second pickup 12 both collect the sound signal of the first user, and the sound signal collected by the first pickup 11 is the first voice information, and the second pickup 12 collects the sound signal. The sound signal is the second voice information, and because the first user is facing the first pickup 11, the voice amplitude of the first voice information collected by the first pickup 11 is greater than the voice of the second voice information collected by the second pickup 12. The amplitude, the controller component 2 compares the voice amplitude of the first voice information with the voice amplitude of the second voice message, and transmits the first voice information with a larger voice amplitude to the main processor component, the main processor The component receives the first voice information sent by the first pickup 11, thereby translating the first voice information from Chinese into English to generate translation information.

Similarly, when the second user speaks, the first pickup 11 and the second pickup 12 both collect the sound signal of the second user, and the sound signal collected by the first pickup 11 is the first voice information, and the second pickup 12 The collected sound signal is the second voice information, and the second user is facing the second pickup 12, so that the voice amplitude of the first voice information collected by the first pickup 11 is smaller than the second voice collected by the second pickup 12. The voice amplitude of the information, the controller component 2 compares the voice amplitude of the first voice information with the voice amplitude of the second voice information, and sends the second voice information with a larger voice amplitude to the main processor component. The main processor component receives the second voice information sent by the second pickup 11, so that the second voice information is translated from English into Chinese to generate translation information, and the user language can be automatically recognized when the first user and the second user perform dialogue communication. And the language that needs to be translated and translated, without the need for the first user and the second user to take the hand and hold the translation key of each party to speak, reducing the user's For improving the user experience.

In this embodiment, the first pickup 11 and the second pickup 12 are installed in the reverse direction by the pickup assembly 1, thereby respectively acquiring the first voice information and the second voice information, when the first user faces the first pickup 11 and the second user When the second pickup 12 is being spoken, the voice amplitude of the first user's voice signal collected by the first pickup 11 is greater than the voice amplitude of the first user's voice signal collected by the second pickup 12, similarly, The voice amplitude of the second user's voice signal collected by the second pickup 12 is greater than the voice amplitude of the second user's voice signal collected by the first pickup 11, and the controller component 2 determines the first voice information and the second voice information. The magnitude of the voice amplitude, thereby determining whether the first user or the second user is speaking. If the first user is speaking, the voice amplitude of the first voice information is greater than the voice amplitude of the second voice information. At this time, the control The component 2 controls the pickup component 1 to transmit the first voice information to the main processor component 3, and the main processor component 3 receives the first voice information and the first voice information is Translating a language into a second language to generate translation information; if the second user is speaking, the speech amplitude of the second speech information is greater than the speech amplitude of the first speech information, and at this time, the controller component 2 controls the pickup component 1 to be the second The voice information is sent to the main processor component 3, and the main processor component 3 receives the second voice information and translates the second voice information from the second language into the first language to generate translation information, where the first language and the second language respectively correspond to The first voice information and the second voice information are automatically translated according to the size of the sound received by the two-way pickup, which reduces the steps of manual setting and improves the user experience.

In an optional embodiment, the automatic translation device of the present invention further includes:

The translation output component 4 is connected to the main processor component 3 for outputting translation information, and the translation information includes at least one of translating speech information and translating text information.

In implementation, the translation output component 4 includes a speaker, and the translated voice information is played through the speaker, wherein the first language of the first user is Japanese and the second language of the second user is Malay, and the first user is facing the first When the pickup 11 speaks, the controller component 2 controls the pickup component 1 to output the first voice information to the main processor component 3, and the main processor component 3 translates the first voice information from Japanese into Malay to generate translation information, and finally passes through the speaker. Playing the translated voice information in the translation information, when the second user is speaking to the second pickup 12, the controller component 2 controls the pickup component 1 to output the second voice information to the main processor component 3, and the main processor component 3 will be the second The voice information is translated into Japanese by Malay to generate translation information, and finally the translated voice information in the translation information is played through the speaker, and the second user can communicate with the first user.

Of course, the translation output component 4 can also be designed with a display screen, and the display screen can display translated text information. When the speaker fails and does not work, the first user and the second user can also communicate through the display screen to improve the user experience.

In an alternative embodiment, the first pickup 11 and the second pickup 12 employ a unidirectional pickup.

The pickup is also called the monitor head. The monitor pickup is a device for collecting the sound of the live environment and then transmitting it to the back-end device. It is composed of a microphone (microphone) and an audio amplifier circuit. Pickups are generally classified into digital pickups and analog pickups. Sound pickups are sound sensing devices that convert analog audio signals into digital signals and perform corresponding digital signal processing through a digital signal processing system.

In implementation, both the first pickup 11 and the second pickup 12 employ a unidirectional pickup, and the first pickup 11 and the second pickup 12 are mounted opposite each other with the highest sensitivity in the direct direction of the unidirectional pickup, and single pointing The pickup has the lowest sensitivity in the back direction. Specifically, the unidirectional pickup receives a positively opposite sound signal that is greater than the sound signal that faces away from the unidirectional pickup, so that the first user faces the first pickup 11 The two users are facing the second pickup 11 as an example. At this time, the first user faces the second pickup 12 and the second user faces the first pickup 11. If the first user speaks, the first pickup 11 and the second pickup 12 are both The voice information of the first user can be received, but the voice amplitude of the first user's voice information received by the first user is greater than the voice amplitude of the first user's voice information received by the second camera, and the controller component 2 The first voice information collected by the first pickup 11 is compared with the voice amplitude of the second voice information collected by the second pickup 12, thereby controlling the one-way pickup with a large voice amplitude. The voice information is output to the main processor component 3, and the main processor component determines that the voice information is sent from the pickup device, thereby determining the language corresponding to the voice information and the language to be translated, thereby automatically recognizing the language and automatically translating the language. Function, saving user's operation process and convenient for users.

In an alternative embodiment, the pickup assembly 1 further comprises a first amplification filtering unit 13 and a second amplification filtering unit 14, the first amplification filtering unit 13 being connected to the output of the first pickup 11 for receiving the first voice information And outputting the first voice information through the amplification filtering process; the second amplification filtering unit 14 is connected to the output end of the second pickup 12 for receiving the second voice information, and after the second voice information is subjected to the amplification filtering process Make the output.

In this embodiment, by providing the first amplification filtering unit 13 and the second amplification filtering unit 14, the first speech information and the second speech information collected by the first pickup 11 and the second pickup 12 are respectively subjected to amplification filtering processing, and finally The controller component 2 outputs the control to the main processor component 3, specifically, when the second user speaks, the second voice information collected by the second sounder 12 has a larger voice amplitude than the first voice collected by the first pickup device 11. After the voice amplitude of the information, the controller component 2 compares the voice amplitudes of the first voice information and the second voice information, and controls the pickup component 1 to output the second voice information to the main processor component 3 for translation to generate translation information, and finally The translation output component 4 performs playback to realize automatic language recognition and translation functions, and the voice information can effectively remove noise and interference information in the voice information after being amplified and filtered, thereby improving the accuracy of voice recognition and language recognition.

The analog-to-digital conversion component 5 is connected to the first amplification filtering unit 13 and the second amplification filtering unit 14 for receiving the first voice information and the second voice information that are output after the amplification filtering process, and respectively respectively, the first voice information And converting the second voice information into the first digital voice information and the second digital voice information; the analog to digital conversion component 4 is further coupled to the controller component 2 and the main processor component 3 for receiving and controlling according to the output of the controller component 2 The signal outputs first digital voice information or second digital voice information to the main processor component 3.

In implementation, the pickup amplifies the sound collected by the microphone through a general analog circuit, and then performs analog-to-digital conversion by the analog-to-digital conversion component 5 and outputs the result to the main processor 3 for translation, which can effectively improve the accuracy of identifying the voice information, and improve the product. quality.

In one embodiment, the analog to digital conversion component 5 includes a first analog to digital conversion unit 51 and a second analog to digital conversion unit 52. The first analog to digital conversion unit 51 is coupled to the first amplification filtering unit 13 for receiving amplification filtering. Processing, outputting the first voice information, and converting the first voice information into the first digital voice information; the second analog-to-digital conversion unit 52 is connected to the second amplification filtering unit 14 for receiving the output after the amplification filtering process The second voice information and convert the second voice information into the second digital voice information.

Specifically, the first analog-to-digital conversion unit 51 is connected to the first amplification filtering unit 13 and the controller component 2, and the second analog-to-digital conversion unit 52 is connected to the second amplification filtering unit 14 and the main controller component 2, and the controller component 2 After receiving the first voice information and the second voice information, performing voice amplitude comparison on the first voice information and the second voice information, determining which voice information has a larger voice amplitude signal, and then controlling the analog-to-digital conversion component 5 to output After the analog-to-digital conversion, the voice information with a large amplitude of the voice is sent to the main processor component 3, and finally the main processor component 3 translates and generates the translation information.

Specifically, when the first user is speaking to the first pickup 11, the first voice information collected by the first pickup 11 is much larger than the voice signal of the second voice information collected by the second pickup 12, and the first voice information is passed through An amplification filtering processing unit 13 performs amplification filtering processing, and then transmits to the first analog-to-digital conversion unit 51 to convert the first voice information after the amplification filtering processing into the first digital voice information, and the second voice information passes the second amplification filtering. The processing unit 14 performs amplification filtering processing, and then sends the second voice information to the second digital voice information after being subjected to the amplification filtering process, and the amplified and filtered voice information is performed by the controller component 2; Comparing the amplitude of the voice, determining that the signal of the voice information of the first voice information and the second voice information is larger, thereby determining that the signal of the digital voice information of the one road is larger, and then outputting the digital voice information with a larger control signal to the a main processor component 3, the main processor component 3 is based on an analog to digital conversion unit that outputs the digital voice information, The language corresponding to the digital voice information and the language to be translated may be known; the language corresponding to the first user is X and the language corresponding to the second user is Y, and the user may operate the corresponding APP application corresponding to the first selection. The pickup 11 corresponds to the language X and the second pickup 12 corresponds to the language Y. The system automatically sets the language of the translation to X and Y. Since the first user is facing the first pickup 11, when the first user speaks, the first user A voice message is sent to the main processor component 3, and the main processor component 3 can translate the first voice information from X to Y. Similarly, when the second user speaks, the main processor component 3 can The second voice information is translated from Y to X, which can automatically recognize the user's language and automatically translate.

Of course, the language corresponding to the first user is X and the language corresponding to the second user is Y. The first user is speaking the first language wake-up word to the first pickup 11 and the second user is talking to the second pickup 12 The second language wake-up word, the main processor component 3 can determine the first user corresponding language X according to the first language wake-up word and the second language wake-up word, and the second user corresponding language Y, the system automatically sets the translated language For X and Y translation, since the first user is facing the first pickup 11, when the first user speaks, the first voice information is sent to the main processor component 3, and the main processor component 3 can be the first The voice information is translated from X to Y. Similarly, when the second user speaks, the main processor component 3 can translate the second voice information from Y to X, can automatically recognize the user's language and automatically translate.

In an alternative embodiment, the controller component 2 comprises a comparison unit 21 and a control unit 22 connected to the comparison unit 21, the comparison unit 21 being connected to the first amplification filtering unit 13 and the second amplification filtering unit 14, for comparing a voice amplitude of the voice information and the second voice information and outputting a comparison signal to the control unit 22; the control unit 22 is coupled to the first analog to digital conversion unit 51 and the second analog to digital conversion unit 52 for controlling the first signal according to the comparison signal An analog-to-digital conversion unit 51 and a second analog-to-digital conversion unit 52 output digital voice information having a large amplitude of speech to the main processor unit 3.

In implementation, when the automatic translation apparatus of the present invention is placed between the first user and the second user, the first pickup 11 and the second pickup 12 are respectively facing the first user and the second user, and the first user is facing the first pickup 11 speaking the language wake-up words corresponding to the first language, the first pickup 11 and the second pickup 12 both receive the language wake-up words, but the signal of the first voice information collected by the first pickup 11 is larger, thereby determining The first pickup 11 corresponds to the first language. Similarly, the second user is speaking the second wake-up word corresponding to the second language, and the first pickup 11 and the second pickup 12 both receive the voice wake-up. The word, but because the signal of the second voice information collected by the second pickup 12 is large, it is determined that the second pickup 12 corresponds to the second language.

In a specific implementation, the silence detection threshold may also be set in the system. When the voice signal amplitude of the first voice information of the first user is detected to exceed the silence detection threshold, if the system recognizes that the first voice information is in the first language, In the language awakening word, it is determined that the first voice information having a larger signal corresponds to the first language, and similarly, when the voice signal amplitude of the second voice information exceeds the silence detection threshold, if the system recognizes that the second voice information is the second In the language of the language, the second speech information corresponding to the signal is determined to correspond to the second language, and the system automatically sets the language of the translation as the first language and the second language, and the first voice information is The first language is translated into the second language, and the second voice information is translated into the first language for the second language, and the corresponding translation function is automatically activated.

When the first user is speaking to the first pickup 11, the first voice information is amplified and filtered by the first amplification filtering processing unit 13 and then sent to the first analog to digital conversion unit 51 for conversion to the first digital voice information, and the second voice. The information is amplified and filtered by the second amplification filtering processing unit 14 and then sent to the second analog-to-digital conversion unit 52 for conversion to the second digital voice information. Meanwhile, the first amplification filtering unit 13 and the second amplification filtering unit 14 are connected to the comparison unit 21. The comparing unit 21 receives the first speech information and the second speech information subjected to the amplification filtering process and compares them to determine which one of the signals is large, and transmits the digital voice information of the larger one of the signals by the control unit 22 to the main unit. Processor component 3 performs the processing.

In a specific implementation, when translating the output, the main processor component 3 can also output a signal to the control unit 22 to control the analog-to-digital conversion component 5 to stop the conversion by the control unit 22, so that the speech and translation outputs are half-duplex, avoiding each other. influences.

In another alternative embodiment, the analog to digital conversion component 5 can also be designed to have only the first analog to digital conversion unit 51, the first analog to digital conversion unit 51 is connected to the control unit 22 and the main processor component 3, and the control unit 22 The first amplification filtering unit 13 and the second amplification filtering unit 14 are connected. After the first speech information and the second speech information are subjected to amplification filtering and amplitude comparison, the control unit 22 controls the first amplification filtering unit 13 and the second amplification filtering unit 14 The voice information with a large amplitude of the voice information is transmitted to the first analog-to-digital conversion unit 51 for analog-to-digital conversion, and the control unit 22 outputs a high-low level signal to the main processor component 3 to notify the main processor component 3 that the main processor component 3 is receiving. Which one is the voice signal, so that the main processor component 3 can determine whether the voice signal is the first voice signal or the second voice signal, thereby realizing automatic language recognition and translation functions, and reducing the cost investment. Of course, the main processor component 3 can also output a signal to the control unit 22, and the first analog-to-digital conversion unit 51 is controlled by the control unit 22 to stop the conversion, so that the speech and translation are output to a half-duplex state, thereby avoiding mutual influence.

On the other hand, as shown in FIG. 5, the present invention further provides an automatic translation method, which is applied to the automatic translation apparatus as described above, and includes:

Step S1, acquiring first voice information collected by the first pickup and second voice information collected by the second pickup;

Step S2, determining whether the voice amplitude of the first voice information is greater than the voice amplitude of the second voice information;

Step S3, if yes, translating the first speech information from the first language to the second language to generate the first translation information, otherwise translating the second speech information from the second language to the first language to generate the second translation information The first language and the second language respectively correspond to the first voice information and the second voice information.

In implementation, the first pickup and the second pickup adopt a unidirectional pickup and are installed opposite to each other, by acquiring the first voice information and the second voice information, and determining whether the voice amplitude of the first voice information is greater than the second voice information. The voice amplitude, if yes, determining that the first user is speaking, because the first user corresponds to the first language and the second user corresponds to the second language, the first voice information can be translated from the first language to the second language. The first translation information, similarly, if the voice amplitude of the first voice information is smaller than the voice amplitude of the second voice information, determining that the second user is speaking, and translating the second voice information from the second language to the first language Generating the second translation information to implement a translation function between the first language and the second language, wherein the first language and the second voice are respectively selected by the first user and the second user through the translation machine.

In this embodiment, the two-way unidirectional pickups respectively receive the voices of both parties of the conversation. When one party speaks, the signal output from the speaker of the speaker is larger than the signal output by the other pickup, and it is easy to distinguish which party is by the comparator. In speech, it is not easy to malfunction, and only the voice information of the speaking party is sent to the main processor component for translation processing. The unidirectional pickup also helps to reduce the influence of surrounding noise, improve the translation accuracy and the translation output effect, and improve the user. Experience.

In one embodiment, before the step of acquiring the first voice information collected by the first pickup and the second voice information collected by the second pickup, the method includes:

Step S4, acquiring first wake-up voice information input by the first user and second wake-up voice information input by the second user;

Step S5: Acquire a first language corresponding to the first wake-up voice information and a second language corresponding to the second wake-up voice information according to the preset voice library.

In the implementation, before the formal conversation conversation, the user can set the respective language by the wake-up words. Specifically, the first user is speaking the language wake-up words of the first language to the first pickup, and the first pickup collects the first user. The first wake-up voice information, and then acquire the first language of the first user according to the first wake-up voice information and the preset voice library. Similarly, the second user is speaking the second-language language wake-up word to the second pickup, The second pickup acquires the second wake-up voice information of the second user, and then acquires the second language of the second user according to the second wake-up voice information and the preset voice library, so that the next communication conversation step can be performed, from the first of the two sides of the talk. By awakening the speech information and the language wake-up words in the second wake-up speech information, the respective languages of the speech and the languages to be translated can be automatically determined, thereby eliminating the cumbersome operation of manually setting the respective languages and key translations, and improving the user experience.

In an optional embodiment, the step of acquiring the first language corresponding to the first wake-up voice information and the second language corresponding to the second wake-up voice information according to the preset voice library includes:

Step S51, determining whether the voice amplitudes of the first wake-up voice information and the second wake-up voice information are both greater than a preset silence detection threshold;

Step S52, if yes, acquiring a second language wake-up word in the first language wake-up word and the second wake-up voice information in the first wake-up voice information;

Step S53: Acquire a first language that matches the first wake-up word and a second language that matches the second wake-up word according to the preset voice library.

In implementation, by setting a preset silence detection (VAD) threshold, when the amplitudes of the first wake-up voice information and the second wake-up voice information respectively collected by the first pickup and the second pickup exceed the preset silence detection When the threshold is used, the system performs the language recognition. In addition, when the amplitudes of the first voice information and the second voice information collected by the first pickup and the second pickup respectively exceed the preset silence detection threshold, the system The translation function will be performed to avoid the misoperation caused by the user's breathing or the surrounding environment, and improve the accuracy of speech recognition and language recognition.

In an optional embodiment, after the step of acquiring the first language matching the first wake-up word and the second language matching the second wake-up word according to the preset voice library, the method further includes:

In step S54, the prompt sound is turned on by playing the translation function in the first language and the second language respectively.

In the implementation, after the first language and the second language are acquired, the automatic translation device reminds the user that the mutual translation function is currently enabled by means of a voice prompt. Specifically, the automatic translation device performs voice through the first language and the second language respectively. The broadcast translation function turns on the prompt tone to inform the user that the conversation can be opened. In another optional embodiment, the automatic translation device may further enable the text display translation function by using the first language and the second language on the display screen to inform the user that the conversation can be opened.

In an optional embodiment, the step of determining whether the voice amplitude of the first voice information is greater than the voice amplitude of the second voice information includes:

Step S21, performing amplification filtering on the first voice information and the second voice information respectively;

Step S22, it is determined whether the amplitude of the voice of the first voice information after the amplification and filtering process is greater than the voice amplitude of the second voice information after the amplification and filtering process.

In an implementation, the automatic translation device receives the first voice information acquired by the first pickup through the first amplification filtering unit, and performs output after the amplification filtering process; and receives the second voice information acquired by the second pickup through the second amplification filtering unit, And after the amplification filtering process, the output is performed; finally, the amplitude of the speech of the first speech information and the second speech information after the amplification filtering process is compared. In some embodiments, after comparison, a specific comparison signal is obtained, for the control unit to control, according to the comparison signal, the output digital speech with a large amplitude of the speech in the first analog to digital conversion unit and the second mode conversion unit according to the comparison signal. Information to the main processor component. After the amplification and filtering process, the voice information can effectively remove the noise and interference information in the voice information, and improve the accuracy of voice recognition and language recognition.

In an optional embodiment, after the step of performing the amplification filtering process on the first voice information and the second voice information respectively, the method further includes:

Step S23, converting the first voice information after the amplification and filtering processing into the first digital voice information and converting the second voice information after the amplification and filtering processing into the second digital voice information.

In an implementation, the automatic translation device converts the first voice information after the amplification and filtering processing into the first digital voice information, and converts the second voice information after the amplification and filtering processing into the second digital voice information, in some embodiments. The first digital voice information and the second digital voice information are both digital signals. In a specific application, the pickup amplifies the voice information collected by the microphone through a general analog circuit, and then performs analog-to-digital conversion of the voice information through the analog-to-digital conversion component, and then outputs the digital voice information to the main processor for translation, which can effectively improve the recognition voice information. Accuracy and improve product quality.

In an optional embodiment, if yes, the first voice information is translated from the first language to the second language to generate the first translation information, otherwise the second voice information is translated from the second language into the first language. The step of generating the second translation information, wherein the first language and the second language respectively correspond to the first voice information and the second voice information, including:

Step S31, if the voice amplitude of the first voice information is greater than the voice amplitude of the second voice information, translating the first digital voice information from the first language to the second language generation Decoding the first translation information, otherwise translating the second digital speech information from the second language into the first language to generate the second translation information, wherein the first language and the second language respectively Corresponding to the first digital voice information and the second digital voice information.

In the implementation, specifically, the comparison signal determines that the signal of the voice information of the first voice information and the second voice information is larger, thereby determining that the signal of the digital voice information of the one channel is larger, and then the control signal is larger. The digital voice information is output, and the main processor component can know the language corresponding to the digital voice information and the language to be translated according to the analog-to-digital conversion unit that outputs the digital voice information.

In an optional embodiment, if yes, the first voice information is translated from the first language to the second language to generate the first translation information, otherwise the second voice information is translated from the second language to the first language generation The second translation information, wherein the first language and the second language respectively correspond to the first voice information and the second voice information, the method further includes:

In step S6, the first translation information or the second translation information is outputted in a manner of translating the voice information and/or translating the text information.

In implementation, the automatic translation device converts the first voice information from the first language into the second language to generate the first translation information and the second voice information from the second language to the first language to generate the second translation information, and finally passes The speaker plays the translated voice information or displays the translated text information through the display screen, so that the second user can communicate with the first user, and the translation information is output in various ways. When the speaker fails and does not work, the first user and the second user You can also communicate through the display to improve the user experience.

As shown in FIG. 6, the present invention further provides a computer device including a memory 1003 and a processor 1002. The memory 1003 stores a computer program 1004. When the processor 1002 executes the computer program 1004, the steps of any of the above methods are implemented. The method includes: acquiring first voice information collected by the first pickup and second voice information collected by the second sounder; determining whether a voice amplitude of the first voice information is greater than a voice amplitude of the second voice information; if yes, using the first voice The information is translated from the first language into the second language to generate the first translation information, otherwise the second voice information is translated from the second language into the first language to generate the second translation information, wherein the first language and the second language respectively correspond to the first language Voice information and second voice information.

Claims

An automatic translation device, comprising:

a pickup assembly comprising a first mounted pickup and a second pickup, the first pickup for collecting first voice information, and the second pickup for collecting second voice information;

a controller component, coupled to the first pickup and the second pickup, for receiving and comparing a magnitude of a voice amplitude of the first voice information and the second voice information, and controlling the sound output of the pickup component a voice information with a large amplitude;

a main processor component, coupled to the pickup component, configured to receive a voice information of a larger amplitude of the voice output by the pickup component, and to the voice according to the predetermined first language and the second language The voice information of the larger amplitude is translated to generate translation information, and the first language and the second language respectively correspond to the first voice information and the second voice information.
The automatic translation apparatus according to claim 1, further comprising:

A translation output component coupled to the main processor component for outputting the translation information.
The automatic translation apparatus according to claim 2, wherein said translation output component comprises a speaker for outputting translated voice information in said translation information.
The automatic translation apparatus according to claim 2, wherein said translation output component comprises a display screen for outputting translated text information in said translation information.
The automatic translation apparatus according to claim 1, wherein said first pickup and said second pickup employ a unidirectional pickup.
The automatic translation apparatus according to any one of claims 1 to 5, wherein the pickup assembly further includes a first amplification filtering unit and a second amplification filtering unit, wherein the first amplification filtering unit is connected to the first An output end of a pickup for receiving the first voice information and outputting after being subjected to amplification filtering processing; the second amplification filtering unit is connected to an output end of the second pickup for receiving the second voice The information is output after being amplified and filtered.
The automatic translation apparatus according to claim 6, further comprising:

An analog-to-digital conversion component is connected to the first amplification filtering unit and the second amplification filtering unit, and configured to receive the first voice information and the second voice information that are output after being subjected to amplification filtering processing, and respectively Converting the first voice information and the second voice information into first digital voice information and second digital voice information; the analog to digital conversion component is further connected to the controller component and the main processor component, Receiving and outputting the first digital voice information or the second digital voice information to the main processor component according to a control signal output by the controller component.
The automatic translation apparatus according to claim 7, wherein said analog to digital conversion unit comprises a first analog to digital conversion unit and a second analog to digital conversion unit, said first analog to digital conversion unit and said first amplification a filtering unit connection, configured to receive the first voice information outputted after being subjected to amplification filtering processing, and converted into first digital voice information; the second analog-to-digital conversion unit is connected to the second amplification filtering unit, and configured to The second voice information outputted after being subjected to the amplification filtering process is received and converted into second digital voice information.
The automatic translation apparatus according to claim 8, wherein said controller component comprises a comparison unit and a control unit connected to said comparison unit, said comparison unit and said first amplification filter unit and said a second amplification filtering unit connection, configured to compare a magnitude of the voice amplitude of the first voice information and the second voice information, and output a comparison signal; the control unit and the first analog to digital conversion unit and the second modulus And a conversion unit connection, configured to control, according to the comparison signal, output digital voice information having a large voice amplitude in the first analog-to-digital conversion unit and the second mode conversion unit to the main processor component.
An automatic translation method for use in an automatic translation apparatus according to any one of claims 1 to 9, characterized in that it comprises:

Acquiring the first voice information collected by the first pickup and the second voice information collected by the second pickup;

Determining whether a voice amplitude of the first voice information is greater than a voice amplitude of the second voice information;

If yes, translating the first voice information from the first language to the second language to generate first translation information, otherwise translating the second voice information from the second language to the first a language to generate second translation information, wherein the first language and the second language respectively correspond to the first voice information and the second voice information.
The automatic translation method according to claim 10, wherein the step of acquiring the first voice information collected by the first pickup and the second voice information collected by the second pickup includes :

Obtaining first wake-up voice information input by the first user and second wake-up voice information input by the second user;

Acquiring the first language corresponding to the first wake-up voice information and the second language corresponding to the second wake-up voice information according to a preset voice library.
The automatic translation method according to claim 11, wherein the acquiring the first language corresponding to the first wake-up voice information and the first corresponding to the second wake-up voice information according to a preset voice library Second language steps, including:

Determining whether the voice amplitudes of the first wake-up voice information and the second wake-up voice information are greater than a preset silence detection threshold;

If yes, acquiring a first language wake-up word in the first wake-up voice information and a second language wake-up word in the second wake-up voice information;

And acquiring, according to the preset voice library, the first language that matches the first wake-up word and the second language that matches the second wake-up word.
The automatic translation method according to claim 12, wherein the first language matching the first wake-up word and the second wake-up word are matched according to the preset voice library After the step of the second language, the method further includes:

The prompt sound is turned on by the voice play translation function corresponding to the first language and the second language, respectively.
The automatic translation method according to claim 10, wherein the step of determining whether the voice amplitude of the first voice information is greater than the voice amplitude of the second voice information comprises:

Performing amplification filtering processing on the first voice information and the second voice information respectively;

And determining whether the voice amplitude corresponding to the first voice information after the amplification filtering process is greater than the voice amplitude corresponding to the second voice information after the amplification filtering process.
The automatic translation method according to claim 14, wherein after the step of performing the amplification and filtering processing on the first voice information and the second voice information, the method further comprises:

Converting the first voice information after the amplification and filtering processing into first digital voice information, and converting the second voice information after the amplification and filtering processing into the second digital voice information.
The automatic translation method according to claim 15, wherein if yes, the first voice information is translated from the first language to the second language to generate first translation information, otherwise the Translating, by the second language, the second language information into the first language to generate second translation information, wherein the first language and the second language respectively correspond to the first voice information and the second voice The steps of the information, including:

If the voice amplitude of the first voice information is greater than the voice amplitude of the second voice information, translating the first digital voice information from the first language to the second language to generate the First translating information, otherwise translating the second digital speech information from the second language to the first language to generate the second translation information, wherein the first language and the second language Corresponding to the first digital voice information and the second digital voice information, respectively.
The automatic translation method according to claim 10, wherein if the first voice information is translated from the first language to the second language, the first translation information is generated, otherwise Translating, by the second language, the second language information into the first language to generate second translation information, wherein the first language and the second language respectively correspond to the first voice information and the second language After the steps of voice information, it also includes:

The first translation information or the second translation information is output by outputting the voice information and/or the translated text information.
A computer device, comprising: a memory, a processor, and a computer program stored on the memory and operable on the processor, the processor executing the computer program as claimed in claim 10 17. The method of any of the preceding claims.