KR102023356B1 - Wearable sign language translation device - Google Patents

Abstract

The present invention relates to a sign language interpreter, and more particularly, to interpret a voice as a sign language and to interpret a sign language as a voice or a text, thereby eliminating the inconvenience of communication between the sign language and the general public, and artificially converting a voice or a sign language to an artificial intelligence. By the algorithm to improve the accuracy of the interpretation, and wearable by a small number of the receiver relates to a wearable sign language interpreter for interpreting a large number of the general public.

Description

Wearable Sign Language Interpreter {WEARABLE SIGN LANGUAGE TRANSLATION DEVICE}

The present invention relates to a sign language interpreter, and more particularly, a wearable sign language interpreter capable of interpreting to a large number of ordinary people by wearing a handful of interpreters by performing an interpreter through artificial intelligence using a sign language interpreter. It is about.

Hearing impairment impairs the ability of hearing, which has a decisive impact on physical, emotional, and intellectual development, impairs the development of language skills, and as a result lacks the concept of language, which benefits education, culture, and diverse information equally to the normal population. It is a difficult reality to receive. In addition, the reality is that people are suffering from "quality of life."

Therefore, people with hearing impairments use sign language a lot to communicate. Generally speaking, a sign language is a non-speech language that is expressed by hand, unlike a speech language. In contrast, a sign language is visually understood and expressed by hand gestures. to be.

However, since the sign language requires a separate learning to be used by a person without general hearing impairment (hereinafter, the general public), it is inefficient because it requires a considerable effort to use it by the general public. Therefore, natural communication between the sign language user (hereinafter referred to as the sign language) and the general public requires considerable effort. Accordingly, the general public and the sign language user may have a method of communicating each other with a letter, but this has caused inconvenience to write a letter to both the sign language user and the general hearing person.

In addition, a method of interpreting voice or sign language in real time has been needed to solve the inconvenience of communication between the listener and the general public, and when a voice or sign language is translated through a machine, a person inputs the voice or sign language. There is a difference in input for each person, resulting in a difference in interpretation.

In order to solve this inconvenience, a commonly used sign language interpreter uses a method of inputting a voice or text of a general person to deliver a interpreted sign language to a callee. Discomfort occurred that the general public should have a sign language interpreter.

Accordingly, a bilingual interpretation is possible between the callee and the general public, and there is no need for an interpretation difference by correcting the difference according to the inputting person. This situation is emerging.

Patent Registration No. 10-0698942 (2007.03.16.)

The technical problem to be achieved by the present invention is a bi-directional interpretation between the receiver and the general public, the difference in interpretation by correcting the difference according to the input does not occur, wearable to perform the interpretation to a large number of the general public wearing a few people To provide a sign language interpreter.

The technical problem to be achieved by the present invention is not limited to the technical problem mentioned above, and other technical problems not mentioned above may be clearly understood by those skilled in the art from the following description. There will be.

In order to achieve the above technical problem, the wearable sign language interpretation apparatus according to the present invention, the body portion provided to be mounted on the user's body, the microphone unit for recognizing the voice generated from the outside through the microphone connected to the body portion, the body portion An input device unit for recognizing text through an input device connected to the voice unit, a voice text analysis unit analyzing the voice or text recognized by the microphone unit or the input unit through artificial intelligence, and the voice or text analyzed by the voice text analysis unit A sign language output unit for outputting a sign language corresponding to a voice or text based on the signal; a camera unit for detecting a user's hand movement through a camera provided in the body unit and recognizing a detected sign language; and a sign language recognized by the camera unit Sign language analysis unit for analyzing the artificial intelligence, based on the sign language analyzed by the sign language analysis unit A voice text output unit for outputting a voice or text corresponding to a sign language, a speaker unit for outputting a voice corresponding to a sign language from the voice text output unit to a speaker, and a text corresponding to a sign language from the voice text output unit to a monitor It provides a monitor unit.

In an exemplary embodiment of the present invention, the voice text analysis unit converts the artificial intelligence to a noise canceling unit for removing noise included in the voice recognized through the microphone unit and a sign language corresponding to the voice from which the noise is removed through the noise removing unit. It may include a first voice algorithm unit for algorithmizing the voice through the first text algorithm unit for algorithmic text through the artificial intelligence to convert to a sign language corresponding to the text input through the input device unit.

In an embodiment of the present invention, the sign language output unit may be configured to artificially encode a sign language corresponding to a speech or text based on the speech or text algorithmized through the first speech algorithm unit and the first text algorithm unit. A sign language algorithm unit may be provided, and the sign language output unit may output the sign language to the monitor to transmit the sign language algorithmized through the first sign language algorithm unit.

In an embodiment of the present invention, the sign language analyzer may be provided with a second sign language algorithm for algorithmic sign language through artificial intelligence to convert the sign language to a voice or text corresponding to a sign language detected by a camera.

In an embodiment of the present disclosure, the voice text output unit may include a second voice algorithm unit and the second voice algorithm that algorithmically encodes a voice corresponding to a sign language content based on a sign language algorithmized through the second sign language algorithm unit. It may include a second text algorithm unit for algorithmicizing the text corresponding to the sign language content through artificial intelligence based on the sign language algorithm through the sign language algorithm unit.

In an embodiment of the present invention, the second voice algorithm unit may output the algorithmized voice through the speaker unit, and the second text algorithm unit may output the algorithmized text through the monitor unit.

In an embodiment of the present invention, the artificial intelligence used in the wearable sign language interpreter may be provided with a deep learning unit to group and classify sign language, voice, or text through deep learning and algorithmize.

In an embodiment of the present invention, a sign language interpretation glasses may be provided to recognize a sign language and perform an interpretation through a wearable sign language interpreter.

In an embodiment of the present invention, a sign language interpretation pendant may be provided to recognize a sign language and perform an interpretation through a wearable sign language interpreter.

According to an embodiment of the present invention, the wearable sign language interpreter interprets the voice as a sign language, and interprets the sign language as voice or text, thereby eliminating the inconvenience of communication between the receiver and the general public, and converting the voice or the sign language to the artificial intelligence. By matching by algorithm, it is possible to improve the accuracy of interpretation without correcting the difference by interpreting the difference according to the person who inputs, and it is effective to perform interpretation to a large number of ordinary people by wearing a few people. have.

The effects of the present invention are not limited to the above-described effects, but should be understood to include all the effects deduced from the configuration of the invention described in the detailed description or claims of the present invention.

1 is a schematic diagram of a wearable sign language interpretation apparatus according to an embodiment of the present invention.
2 is a block diagram of a wearable sign language interpreter according to an embodiment of the present invention.
3 is a schematic configuration diagram of a wearable sign language interpreter according to another embodiment of the present invention.
4 is a flowchart of a sign language interpretation method according to an embodiment of the present invention.
5 is a flowchart of a sign language interpretation method according to another embodiment of the present invention.
6 is a block diagram of a sign language interpretation system according to another embodiment of the present invention.

Hereinafter, with reference to the accompanying drawings will be described the present invention. As those skilled in the art would realize, the described embodiments may be modified in various different ways, all without departing from the spirit or scope of the present invention. In the drawings, parts irrelevant to the description are omitted in order to clearly describe the present invention, and like reference numerals designate like parts throughout the specification.

Throughout the specification, when a part is said to be "connected (connected, contacted, coupled)" with another part, it is not only "directly connected" but also "indirectly connected" with another member in between. "Includes the case. In addition, when a part is said to "include" a certain component, this means that it may further include other components, without excluding the other components unless otherwise stated.

The terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention. Singular expressions include plural expressions unless the context clearly indicates otherwise. As used herein, the terms "comprise" or "have" are intended to indicate that there is a feature, number, step, action, component, part, or combination thereof described on the specification, and one or more other features. It is to be understood that the present invention does not exclude the possibility of the presence or the addition of numbers, steps, operations, components, components, or a combination thereof.

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

1 is a schematic configuration diagram of a wearable sign language interpretation apparatus according to an embodiment of the present invention, Figure 2 is a block diagram of a wearable sign language interpretation apparatus according to an embodiment of the present invention.

1 to 2, the wearable sign language interpretation apparatus 200 according to the present embodiment is provided with a body portion to be mounted to the user's body.

More specifically, when the user is a receiver, the body 110 is provided to be worn on a human body such as a necklace, earrings, glasses, and the like. The body portion 110 is provided with a camera portion 111 is equipped with a camera that can detect the user's hand movements.

In addition, the user wears the body portion 110 manufactured in the shape of glasses, and detects the hand movement for sign language through a camera provided in the camera portion 111 mounted on the body portion 110.

When the hand movement is detected, the sign language analyzer 250 of the sign language interpretation system 100 algorithmizes the content of the sign language recognized by the camera 111 and outputs a voice corresponding to the sign language contents. do.

In this case, the speaker corresponding to the sign language content is algorithmized to match the coded sign language content, and the speaker unit 140 having a speaker connected to the wearable sign language interpreter by wire or wirelessly. Will output via

The user refers to a person who uses the wearable sign language interpreter 200, and may be a signer who performs a sign language according to a situation, and may be a general person who inputs a voice or text.

In addition, a monitor unit having a monitor connected to the wearable sign language interpreter 200 and wired or wirelessly by matching the algorithm and the sign language content by algorithmized text corresponding to the sign language content and the matching ( 160).

By matching the algorithmic sign language and the algorithmic speech and text, the sign language is re-recognized to match the speech and text when it is not matched.

On the contrary, when the user is a general person, the user wears the body part 110 manufactured in the shape of glasses, and the microphone is mounted on the body part 110 or is connected to the body part 110 by wire or wireless. The microphone 120 may recognize the voice.

The microphone unit 120 recognizes the user's voice and analyzes the recognized voice through the voice text analyzer 220. Accordingly, when the user recognizes the voice through the microphone unit 120, the voice text analyzer 220 algorithms the voice and algorithms a sign language corresponding to the content of the voice. By matching the algorithmic voice and the algorithmic sign language, the content of the sign language is output through the monitor unit 160 equipped with a monitor.

In addition, the user wears the body part 110 manufactured in the shape of glasses, and recognizes the text through the input device unit 170 having an input device connected to the body part 110 in a wired or wireless manner.

The input device unit 170 recognizes the text input of the user and analyzes the recognized text through the voice text analysis unit 220. Therefore, when the user recognizes the text through the input device, the text analysis unit 220 algorithmizes the text and algorithms a sign language corresponding to the content of the text. By matching the algorithmized text and the algorithmized sign language, the content of the sign language is output through the monitor unit 160 equipped with a monitor.

On the other hand, by matching the algorithmic speech and text with the algorithmic sign language, if it is not matched to recognize the speech and text again to match the sign language.

Referring to FIG. 2, the wearable sign language interpreter 200 according to the present exemplary embodiment includes a voice text recognition unit 210 that recognizes text through an external voice or an input device generated through the microphone unit 110. .

In more detail, a microphone unit 120 configured as a microphone for recognizing a voice is provided, and the microphone is a device that recognizes what the user is talking through voice. The microphone may be mounted on the wearable sign language interpreter 200 or may be used separately to interwork with the wearable sign language interpreter 200. It is also possible to use the microphone through a smart terminal such as a mobile phone. The microphone is a device for converting a commonly used voice or sound into an electric pulse, and a detailed configuration of the microphone will be omitted.

In addition, the input device unit 170 is connected to the wearable sign language interpreter 200 to input text. The input device unit 170 transmits the input text to the wearable sign language interpreter 200 by interworking with the wearable sign language interpreter 200 in a wired or wireless manner such as a keyboard, a screen pad, or a mobile phone.

On the other hand, the speech text analysis unit 220 for analyzing the speech or text recognized from the speech text recognition unit 210 through artificial intelligence is provided.

In more detail, the voice text analysis unit 220 recognizes the voice recognized through the microphone unit 120 or the text input through the input device unit 170 through the voice text recognition unit 210, Analyze this through artificial intelligence. Information such as voice or text input through the microphone or the input device unit 170 is different according to the person expressing, and the artificial intelligence corrects the difference and analyzes information commonly expressed. That is, even if the same expression is different for each person who inputs, the voice text analysis unit 220 may distinguish the intentions of the person who inputs even the difference, so that the wearable sign language interpreter 200 can perform the interpretation. Analyzes the text input through the voice or the input unit 170 recognized through the microphone.

In addition, a sign language output unit 230 is provided to output a sign language through the wearable sign language interpreter 200 based on the speech or text analyzed by the speech text analyzer 220.

More specifically, the sign language output unit 230 may output a sign language corresponding to the voice or text analyzed by the voice text analyzer 220 through artificial intelligence. The wearable sign language interpreter 200 is used to output a sign language content, and the wearable sign language interpreter 200 may be provided in the same shape as a human hand to accurately deliver the sign language content. In addition, it is possible to output such a sign language through a monitor connected to the robot. Therefore, the wearable sign language interpreter 200 may be transmitted to the short distance target, and the sign language may be delivered to the remote target through the monitor.

Meanwhile, the voice text analysis unit 220 includes a noise removing unit 221, a first voice algorithm unit 222, and a first text algorithm unit 223.

More specifically, the noise removing unit 221 is provided to remove noise included in the voice recognized through the microphone. Therefore, it is possible to recognize the voice more accurately through the microphone. In addition, a first voice algorithm unit 222 is provided to algorithmize the voice through artificial intelligence so as to translate the sign language corresponding to the voice from which the noise is removed through the noise removing unit 221. The artificial intelligence applied to the first voice algorithm unit 222 may be variously applied, but preferably, the information recognized through the deep learning unit 270 using deep learning is clustered, classified and algorithmized.

In addition, the first text algorithm unit 223 algorithms the text through artificial intelligence to sign language translation corresponding to the text input through the input device unit 170. The artificial intelligence can be variously applied, but preferably, the deep learning unit 270 clusters and classifies the recognized information and algorithms.

Also, the sign language output unit 230 outputs the sign language algorithmized through the first sign language algorithm unit 231 to the wearable sign language interpreter 200.

More specifically, the first sign language algorithm unit 231 is based on the voice or text algorithmized by the first voice algorithm unit 222 and the first text algorithm unit 223, and the voice or text. The corresponding sign language is algorithmized through the deep learning unit 270 using the deep learning, and is output through the wearable sign language interpreter 200.

That is, the recognized voice or text is algorithmized through the deep learning unit 270 in the first voice algorithm unit 222 and the first text algorithm unit 223, and corresponds to the algorithm or voice. The sign language is derived and algorithmized by the first sign language algorithm 231 through the deep learning unit 270. Accordingly, by algorithmizing a voice or text and algorithmic sign language corresponding thereto, a difference according to the user of the wearable sign language interpreter 200 may be compensated for, and an error in interpretation may be prevented.

Meanwhile, the wearable sign language interpreter 200 is provided with a sign language recognition unit 240 that detects a movement through a camera unit 111 provided with a camera and recognizes a detected sign language.

In more detail, the sign language recognition unit 240 photographs the hand of the talker through the camera unit 111 provided with the camera in the wearable sign language interpreter 200, thereby recognizing the sign language. In addition, the sign language recognition unit 240 performs a sign language while the receiver is wearing a glove equipped with a sensor so that the receiver can recognize the sign language, and the wearable sign language interpreter 200 connected to the glove signal from the sensor It is also possible to recognize sign language.

In addition, a sign language analyzer 250 for analyzing a sign language recognized by the sign language recognition unit 240 through artificial intelligence is provided.

In more detail, the sign language analyzer 250 recognizes a sign language from the sign language recognition unit 240 and analyzes the recognized sign language through artificial intelligence. That is, even if the same sign language is performed for each signer who performs the sign language, the difference occurs due to the position, speed, and habits of the hand. The sign language is analyzed through artificial intelligence to include the difference and to deliver a certain sign language. . Therefore, the sign language content allows accurate and consistent analysis of the sign language content even if a difference occurs for each talker.

On the other hand, the speech text output unit 260 for outputting a speech or text corresponding to the sign language based on the sign language analyzed by the sign language analyzer 250 is provided.

More specifically, the voice text output unit 260 may output voice or text to correspond to a sign language analyzed by the sign language analyzer 250. The voice is output through a microphone mounted on the wearable sign language interpreter 200 or a microphone connected to the wearable sign language interpreter 200, and the text is output through a monitor or an external monitor provided in the wearable sign language interpreter 200. It is possible. In addition, the wearable sign language interpreter 200 may be linked with a mobile phone and output through a speaker or a screen provided in the mobile phone.

In addition, the sign language analyzer 250 is provided with a second sign language algorithm unit 251 for algorithmic sign language through artificial intelligence.

More specifically, the second sign language algorithm unit 251 algorithms a sign language through artificial intelligence to translate into a voice or text corresponding to a sign language detected through a camera. The artificial intelligence can be variously applied, but preferably, the information recognized through the deep learning unit 270 using the deep learning is clustered and classified and algorithmized.

Meanwhile, the voice text output unit 260 outputs the voice or text algorithmized through the second sign language algorithm unit 251 to the wearable sign language interpreter 200, or the second text algorithm unit 261 or the second text algorithm. The unit 262 is provided.

In more detail, a second voice algorithm unit 261 is provided to algorithmize a voice corresponding to a sign language content through artificial intelligence based on a sign language algorithmized through the second sign language algorithm unit 251, and the A second text algorithm unit 262 is provided to algorithmize the text corresponding to the sign language content through artificial intelligence based on the sign language algorithmized through the second sign language algorithm unit 251.

That is, the sign language is algorithmized by the talker, and the voice or text corresponding to the sign language is algorithmically outputted through the wearable sign language interpreter 200. The artificial intelligence can be variously applied, but preferably, the information recognized through the deep learning unit 270 using the deep learning is clustered and classified and algorithmized.

In addition, the deep learning unit 270 may be provided with a matching unit 280 to cluster, classify, and match the algorithm and the algorithmized speech or text.

More specifically, the matching unit 280 is a sign language algorithmized by the first and second sign language algorithms 231 and 251 and the first and second voice algorithm parts 222 and 261 or the first. And clustering, classifying, and matching the speech or text algorithmized through the second text algorithm units 223 and 262, respectively. Therefore, since the sign language and the voice or the text are matched, the sign language and the voice or the text can be interpreted more easily, and the deviation according to the user can be reduced through matching.

The voice text output unit 260 may output voice and text corresponding to a sign language, or may output voice and text simultaneously.

3 is a schematic configuration diagram of a wearable sign language interpreter according to another embodiment of the present invention.

2 to 3, the wearable sign language interpreter 300 according to the present embodiment is provided with a body portion 310 to be mounted to the user's body. The user wears a body part 310 prepared to be used as a pendant on a neck, and detects a hand movement for sign language through a camera provided in the camera part 311 mounted to the body part 310.

The wearable sign language interpreter shown in FIG. 3 is the same as the wearable sign language interpreter of FIG. 1 except that the body 310 is manufactured as a pendant, and thus a detailed description thereof will be omitted.

The body 310 is worn on the user's body to recognize the user's hand movement through the camera, so long as it can easily recognize the hand movement is not limited in the wearing position and shape. Therefore, the body portion 310 can be manufactured in various shapes, preferably made of glasses, necklaces, earrings and the like.

4 is a flowchart of a sign language interpretation method according to an embodiment of the present invention.

1 to 4, the wearable sign language interpreter 200 according to the present exemplary embodiment includes a microphone or an input device provided in a wearable sign language interpreter 200 using a voice or text of a user who uses the wearable sign language interpreter. Recognizing through the step (S410).

More specifically, the user may input a voice through the microphone unit 120 provided with a microphone or recognize the text to the wearable sign language interpreter 200 through the input device unit 170. The microphone unit 120 and the input device unit 170 may be mounted on the wearable sign language interpreter 200 or may be connected to the wearable sign language interpreter 200 in a wired or wireless manner to be executed externally.

In addition, the method may include a step (S420) of algorithmicizing the voice or text recognized through the microphone unit 120 or the input device unit 170 through artificial intelligence.

In more detail, the voice recognized through the microphone unit 120 may be algorithmized through artificial intelligence, or the text recognized through the input device unit 170 may be algorithmized through artificial intelligence, and the contents of the recognized voice may be used. Or converts to a sign language corresponding to the content of the recognized text.

In addition, the method may include algorithmicizing a sign language corresponding to voice or text (S430).

In the step S420 of algorithmicizing the recognized voice or text through artificial intelligence, a sign language corresponding to the content of the recognized voice or the content of the recognized text is derived and the sign language is algorithmized.

In addition, the method may include algorithmicizing a speech or text and algorithmic matching a sign language corresponding thereto (S440).

By voicening the voice recognized through the microphone unit 120 through artificial intelligence, and by matching the sign language corresponding to the sign language, it is possible to standardize a sign language corresponding to the voice and to output a sign language corresponding to the voice. It is possible to reduce the errors generated.

By textifying the text recognized through the input device unit 170 through artificial intelligence, and by matching the sign language corresponding to the sign language, it is possible to standardize the sign language corresponding to the text, the output of the sign language corresponding to the text This can reduce the errors that occur at the time.

On the other hand, the artificial intelligence is applicable to a variety of but preferably algorithms through deep learning to cluster and use the recognized sign language and speech or text.

In addition, if matching is not possible in matching the algorithmized speech or text with the algorithmic sign language, the method may further include re-performing the speech or text (S441).

In addition, the method may include performing a step of matching the algorithmized sign language with the algorithmized sign language and outputting a sign language content through the wearable sign language interpreter 200 when the matching is possible (S450).

More specifically, the voice recognized through the microphone unit 120 is algorithmized, and a sign language corresponding to the contents of the voice is algorithmically matched with each other. At this time, when the contents of the voice and the contents of the sign language match, the contents of the sign language are output through the wearable sign language interpreter 200.

In addition, the text recognized through the input device unit 170 is algorithmized, and the sign language corresponding to the contents of the text is algorithmized to match each other. At this time, when the contents of the text and the contents of the sign language match, the contents of the sign language are output through the wearable sign language interpreter 200.

5 is a flowchart of a sign language interpretation method according to another embodiment of the present invention.

1 to 5, the sign language interpretation method 500 according to the present embodiment recognizes a sign language of a sign language interpreter using the sign language interpretation system through a camera unit 111 provided in a wearable sign language interpreter 200. It includes a step (S510).

More specifically, the user may recognize the sign language in the wearable sign language interpreter 200 through the camera unit 111 provided with a camera. The camera unit 111 may have a structure mounted in the wearable sign language interpreter 200 or may be connected to the wearable sign language interpreter 200 in a wired or wireless manner to be executed externally.

In addition, the method may include a step S520 of algorithmic sign language recognized by the camera through artificial intelligence.

In more detail, the sign language recognized by the camera provided in the camera unit 111 is algorithmized through artificial intelligence, and converted into speech or text corresponding to the contents of the recognized sign language.

In addition, the method may include a step S530 of algorithmic speech or text corresponding to a sign language.

In step S520 of algorithmicizing the recognized sign language through artificial intelligence, a sign language corresponding to the content of the recognized voice or the content of the recognized text is derived, and the sign language is algorithmized.

In addition, the method may include a step of algorithmicizing a speech or text and algorithmic matching a sign language corresponding thereto (S540).

By voicening the voice recognized through the microphone unit 120 through artificial intelligence, and by matching the sign language corresponding to the sign language, it is possible to standardize a sign language corresponding to the voice and to output a sign language corresponding to the voice. It is possible to reduce the errors generated.

That is, the sign language recognized by the camera unit 111 is algorithmized, and the voice or text corresponding to the content of the sign language is algorithmically matched with each other. At this time, when the content of the sign language matches the content of the voice or text, the content of the voice is output through the speaker unit 140 provided in the wearable sign language interpreter 200, or the content of the text is wearable sign language interpreter. The monitor 200 may be output through the monitor 160 connected to the device 200. The speaker unit 140 and the monitor unit 160 may output audio content and text content, respectively, and may be output simultaneously as necessary.

In addition, if matching is not possible in matching the algorithmized sign language with the algorithmic speech or text, the method may further include performing a step of recognizing the sign language (S541).

In addition, performing the step of matching the algorithmic speech or text with the algorithmic sign language, if the match is possible to output the voice through the speaker unit 140 provided in the wearable sign language interpreter 200 or monitor unit 160 It may include the step of outputting the text through (S550).

More specifically, the sign language recognized by the camera unit 111 is algorithmized, and the voice corresponding to the content of the sign language is algorithmically matched with each other. At this time, when the contents of the sign language and the contents of the voice match, the contents of the sign language are output through the speaker unit 140 provided in the wearable sign language interpreter 200.

In addition, the sign language recognized by the camera unit 111 is algorithmized, and the text corresponding to the content of the sign language is algorithmically matched with each other. At this time, when the content of the sign language matches the content of the text, the text is output through the monitor unit 160 connected to the wearable sign language interpreter 200.

Accordingly, since the content of the sign language is output through the speaker unit 140 or the monitor unit 160 only when matching is possible, an error between the content of the sign language and the content of the voice and text may be reduced by correcting according to a difference between users.

6 is a block diagram of a wearable sign language interpreter 700 according to another embodiment of the present invention.

Referring to FIG. 6, the wearable sign language interpreter 700 according to the present invention may include a first user recognition unit 710.

In more detail, the first user recognition unit 710 recognizes a user who uses the wearable sign language interpreter 700 and classifies the user features into a database. The first user recognition unit 710 may recognize the user in various ways. Preferably, the first user recognition unit 710 may detect the user's face through the camera unit 111 or recognize the user's voice through the microphone unit 120. It may be detected and recognized, or may be input by the user through the input device unit 170 to recognize the user.

In addition, the voice text recognition unit 720 for recognizing the voice or text of the user recognized by the first user recognition unit 710 is provided.

In more detail, a microphone unit 120 configured as a microphone for recognizing a voice is provided, and the microphone is a device that recognizes what the user is talking through voice. The microphone may be mounted on the wearable sign language interpreter 700 or may be individually used to be linked wirelessly. It is also possible to use the microphone through a smart terminal such as a mobile phone.

In addition, the input device unit 170 is provided to be connected to the wearable sign language interpreter 700 to input text. The input device unit 170 transfers the input text by interworking with the wearable sign language interpreter 700 in a wired and wireless manner such as a keyboard, a screen pad, or a mobile phone.

Therefore, when the user is recognized through the first user recognition unit 710, and the user recognizes the voice or text by the wearable sign language interpreter 700, the voice or text is recognized by the voice text recognition unit 720. .

On the other hand, the speech or text recognized by the speech text recognition unit 720 is to analyze the speech or text through the speech text analysis unit 730, the speech text analysis unit 730 is a noise removing unit 731, A first voice algorithm portion 732 and a first text algorithm portion 733 are provided. Since the configuration of the noise removing unit 531, the first voice algorithm unit 732, and the first text algorithm unit 733 is the same as that of FIG. 1, a detailed description thereof will be omitted.

When the user recognizes the voice or the text through the voice text recognition unit 720, the first voice algorithm unit 732 and the first text algorithm unit 733 included in the voice text analyzer 730 may algorithmize this. Let's do it. At this time, a voice difference due to a habit or a habit occurs for each user, and a first error detector is provided to detect an error due to the difference.

The first error detection unit 734 is provided to detect an error when an error occurs in the content of the voice or text of the user. Therefore, if an error occurs repeatedly for a specific voice or text, the user is provided to correct it and to translate it into a standardized sign language.

In other words, the user information stored through the first user recognition unit 710 is stored, through which the error of the repetitive voice or text caused by the user's habits or habits such as to interpret the standardized sign language To detect the generated voice or text.

 Therefore, an error of the voice or text detected by the first error detector 734 is converted into a standardized voice or text through the converter 792 included in the deep learning unit 790. As a result, the voice or text converted by the conversion unit 792 is interpreted as a sign language corresponding to this and output through the sign language output unit 740.

For example, when the user recognizes the user through the first user recognition unit 710, the database recognizes the user through the first user recognition unit 710 and recognizes the voice through the voice text recognition unit 720. The voice is analyzed through the voice text analyzer 730. At this time, the first user recognition unit 710 database the habits or habits of the user, when the user habitually recognizes the "A" as "B" to grasp the context of the input voice, etc. error in the voice The first error detection unit 734 determines that there is. Accordingly, the voice having an error is converted into “A” instead of being recognized as “B” input by the user through the conversion unit 792, and finally the sign language is interpreted as a standardized sign language without error. do.

The wearable sign language interpreter according to the present invention may include a second user recognition unit 750.

In more detail, the second user recognition unit 750 recognizes a user who uses the wearable sign language interpreter 700 and classifies the user features into a database. The second user recognition unit 750 may recognize the user in various ways, preferably by detecting the face or sign language of the user through the camera unit 111 provided in the wearable sign language interpreter 700 or The user can recognize the user by inputting it through the input device.

In addition, a sign language recognition unit 760 for recognizing a sign language of a user recognized through the second user recognition unit 750 is provided.

In more detail, a camera unit 111 configured as a camera is provided to recognize a sign language, and the camera is a device that recognizes a user's intention through a sign language input by a user. The camera may be mounted on the wearable sign language interpreter 700 or may be used separately and interlocked wirelessly. It is also possible to be used as a camera through a smart terminal such as a mobile phone.

Therefore, when the user is recognized through the second user recognition unit 750, and the user recognizes the sign language by the wearable sign language interpreter 700, the sign language is recognized by the sign language recognition unit 760.

On the other hand, the sign language recognized by the sign language recognition unit 760 analyzes the sign language through the sign language analyzer 770, and the sign language analyzer 770 includes a second sign language algorithm part 771. Since the configuration of the second sign language algorithm part 771 is the same as that of FIG. 1, a detailed description thereof will be omitted.

When the user recognizes the sign language through the sign language recognition unit 760, the second sign language algorithm part 771 algorithms the sign language. At this time, a difference in sign language due to habits or habits is generated for each user, and a second error detector 772 is provided to detect a sign language error due to the difference.

The second error detection unit 772 is provided to detect an error when an error occurs in the content of the sign language of the user. Therefore, if an error occurs repeatedly for a particular sign language, the user is provided to correct the standardized sign language.

In other words, the user information stored through the second user recognition unit 750 is stored, through which the error of repetitive sign language caused by the user's habits or habits such as to translate the standardized voice or text To detect the generated sign language.

Accordingly, the sign language error detected by the second error detector 772 is converted into standardized sign language through the converter 792 included in the deep learning unit 790. As a result, the sign language converted by the converter 792 is interpreted as a voice or text corresponding thereto and output through the voice text output unit 780.

The foregoing description of the present invention is intended for illustration, and it will be understood by those skilled in the art that the present invention may be easily modified in other specific forms without changing the technical spirit or essential features of the present invention. will be. Therefore, it should be understood that the embodiments described above are exemplary in all respects and not restrictive. For example, each component described as a single type may be implemented in a distributed manner, and similarly, components described as distributed may be implemented in a combined form.

The scope of the present invention is represented by the following claims, and it should be construed that all changes or modifications derived from the meaning and scope of the claims and their equivalents are included in the scope of the present invention.

200, 700: wearable sign language interpreter
110, 310: body part
111, 311: camera section
120, 320: microphone section
140, 340: speaker unit
150, 350: robotic arm
160, 360: monitor
170, 370: input unit

Claims (9)

Body portion provided to be mounted to the user's body;
A microphone unit recognizing an externally generated voice through a microphone connected to the body unit;
An input device unit recognizing text through an input device connected to the body unit;
A voice text analysis unit for analyzing the voice or text recognized by the microphone unit or the input unit through artificial intelligence;
A sign language output unit that outputs a sign language corresponding to the voice or text based on the voice or text analyzed by the voice text analyzer;
A camera unit for detecting a hand movement of the user equipped with the body unit through a camera provided in the body unit and recognizing a detected sign language;
A sign language analyzer for analyzing a sign language recognized by the camera unit through artificial intelligence;
A voice text output unit for outputting a voice or text corresponding to a sign language based on a sign language analyzed by the sign language analyzer;
A speaker unit for outputting a voice corresponding to a sign language from the voice text output unit to a speaker; And
A monitor unit for outputting text corresponding to a sign language from the voice text output unit to a monitor;
The apparatus may be configured to detect and recognize a user's face through the camera unit, to detect and recognize a user's voice through a microphone unit, or to input a user through an input device to recognize the user and classify the user's features into a database. 1 user recognition unit;
If an error occurs in the contents of the voice or text of the user recognized by the first user recognition unit, the error is detected, and if a specific voice or text is repeatedly generated by the user, the error is corrected and standardized sign language A first error detection unit provided to translate into;
A second user recognition unit configured to detect and recognize a face or a sign language of the user through the camera unit, or to recognize the user by inputting the user through an input device, and to classify and classify the user features; And
If an error occurs in the content of the user's sign language recognized by the second user recognition unit, the error is detected, and if a specific sign language is repeatedly generated by the user, the error is corrected and interpreted as a standardized voice or text. A second error detection unit provided to make;
Wearable sign language interpreter comprising a.
The method of claim 1, wherein the voice text analysis unit,
A noise removing unit for removing noise included in the voice recognized by the microphone unit; And
A first speech algorithm unit for algorithmizing the speech through artificial intelligence to convert the speech into a sign language corresponding to the speech from which the noise is removed;
A first text algorithm unit which algorithms the text through artificial intelligence so as to convert a sign language corresponding to the text input through the input device unit;
Wearable sign language interpreter comprising a.
The method of claim 2, wherein the sign language output unit,
A first sign language algorithm is provided to artificially algorithmic a sign language corresponding to a voice or text based on the voice or text algorithmized through the first voice algorithm unit and the first text algorithm unit, and the sign language output unit is provided. Wearable sign language interpreter, characterized in that for outputting to the monitor to pass the algorithm for the sign language through the first sign language algorithm.
The method of claim 1, wherein the sign language analyzer,
The wearable sign language interpreter, characterized in that the second sign language algorithm for converting the sign language through the artificial intelligence to convert the speech or text corresponding to the sign language detected by the camera.
The method of claim 4, wherein the voice text output unit,
A second voice algorithm unit which algorithms a voice corresponding to a sign language content through artificial intelligence based on a sign language algorithmized through the second sign language algorithm unit; And
A second text algorithm unit which algorithms the text corresponding to the sign language content through artificial intelligence based on the sign language algorithmized through the second sign language algorithm unit;
Wearable sign language interpreter comprising a.
The wearable sign language interpreter of claim 5, wherein the second voice algorithm unit outputs an algorithmized voice through the speaker unit, and the second text algorithm unit outputs an algorithmized text through the monitor unit.
The wearable sign language interpreter according to claim 1, wherein the artificial intelligence used in the wearable sign language interpreter is provided with a deep learning unit to group, classify, and classify sign language, voice, or text through deep learning.
The sign language interpreter glasses according to any one of claims 1 to 7, wherein the wearable sign language interpreter is provided in front of a user's eyes to recognize the sign language and perform an interpreter.
A sign language interpreter pendant according to any one of claims 1 to 7, wherein the wearable sign language interpreter is provided to recognize a sign language and perform an interpreter on a user's neck or ear.

Images