KR102613249B1 - Method for providing security and personalization service through personal voice identification based on artificial intelligence and apparatus therefor - Google Patents

Abstract

The present invention relates to a method of providing security and personalization services through artificial intelligence-based personal voice identification and a device therefor. According to one aspect, a method of providing security and personalization services in a set-top box paired with a remote control includes information on each profile from the remote control. Receiving a first audio file and registering a security authentication audio file for each profile; Receiving a second audio file from the remote control; and a speaker based on the security authentication audio file registered for each profile and the second audio file. It may include the steps of authenticating and identifying, generating a personalized user interface screen based on the speaker identification result, and transmitting the generated personalized user interface screen to a TV. Therefore, the present invention has the advantage of providing enhanced security and personalized services using an artificial intelligence-based set-top box.

Description

{Method for providing security and personalization service through personal voice identification based on artificial intelligence and apparatus therefor}

The present invention relates to an artificial intelligence set-top box, and more specifically, to technology that provides security and personalized services through personal voice identification based on artificial intelligence mounted on a device.

Set-Top Box (STB) literally means a box placed on the floor and generally refers to a device for receiving digital broadcasts.

A set-top box is a device that receives terrestrial broadcasting, cable broadcasting, and satellite broadcasting at home and watches them on a connected TV. Recently, it is a device that connects a set-top box and a phone line (or dedicated line) to watch broadcast programs on TV as well as view/select/watch various contents such as YouTube, movies, music, shopping, education, and games through Internet access. It is being utilized.

The digital broadcasting server compresses video, voice, and data signals into multiplexes and then modulates them and transmits them. The set-top box receives signals transmitted by a digital broadcasting server through a digital tuner and then decompresses and separates each signal through a demodulation process.

Recently, set-top boxes equipped with smart speakers (i.e., artificial intelligence speakers) have been released mainly by telecommunication companies. The set-top boxes recognize the user's voice input through the speaker through linkage with a dedicated voice recognition server, and the recognition results Based on this, it was possible to provide broadcast channel search services, music playback services, various content search services, Internet information search services, and lighting and home appliance control services.

However, as shown in Figure 1, the security technology applied to currently commercialized set-top boxes sets a numeric profile lock password through a remote control, and unlocks the user profile by entering the set password. It has been done. Therefore, passwords can be easily exposed to other people, and there is a problem with security vulnerabilities.

In addition, the voice recognition function installed in conventional set-top boxes had the disadvantage of not being able to provide personalized services because it was not equipped with a separate speaker identification function.

The purpose of the present invention is to provide a method and device for providing security and personalized services through artificial intelligence-based personal voice identification.

Another object of the present invention is to provide an on-device voice recognition-based set-top that can provide security and personalized services by directly recognizing and analyzing the user's voice inside the set-top box without linking with an external server for voice recognition. The goal is to provide a method of providing box security and personalized services.

Another object of the present invention is to provide a set-top box equipped with an on-device voice recognition function that enables quick profile switching by identifying the speaker through the voice recognition function installed inside the set-top box without linking with a separate server. is to provide.

Another purpose of the present invention is to enhance the security of personal profiles compared to the conventional password input method by providing functions such as profile lock setting/unlocking and profile switching through on-device-based voice recognition. In addition, it provides a method and device for providing set-top box security and personalization services based on on-device voice recognition that can quickly provide customized personal services to customers according to profile switching.

The technical problems of the present invention are not limited to the technical problems mentioned above, and other technical problems not mentioned will be clearly understood by those skilled in the art from the description below.

According to one aspect, a method of providing security and personalization services in a set-top box paired with a remote control includes receiving a first audio file for each profile from the remote control and registering a security authentication audio file for each profile, and receiving a second audio file from the remote control. receiving, authenticating and identifying a speaker based on the security authentication audio file registered for each profile and the second audio file, and generating a personalized user interface screen based on the speaker identification result. It may include transmitting the personalized user interface screen to the TV.

In an embodiment, the first audio file may include a specific predefined phrase or lock keyword.

In an embodiment, the first audio file and the second audio file may be created by the remote control in response to a user-selected profile on the home screen of the TV.

In an embodiment, the microphone of the remote control is activated based on the first button provided on the remote control being pressed for a first time while the profile lock setting screen is displayed on the TV, and the microphone of the remote control is activated based on the voice recorded through the activated microphone. Thus, the first audio file can be created.

In an embodiment, while the profile unlock screen is displayed on the TV, the microphone of the remote control is activated based on the first button provided on the remote control being pressed for a second time, and the voice recorded through the activated microphone is activated. Based on this, the second audio file may be created.

In an embodiment, the step of authenticating and identifying a speaker based on the security authentication audio file registered for each profile and the second audio file includes phonological analysis for each of the security authentication audio file registered for each profile and the second audio file. It may include the step of authenticating and identifying the speaker corresponding to the second audio file based on the step of performing and the result of the phoneme analysis.

In an embodiment, the phonological analysis may be performed using a phonological analysis engine mounted on the set-top box without linking with an external server.

In an embodiment, the step of authenticating and identifying a speaker based on the security authentication audio file registered for each profile and the second audio file includes preprocessing each of the security authentication audio file registered for each profile and the second audio file. It may include extracting a voice feature vector and performing voice recognition machine learning based on the extracted voice feature vector to authenticate and identify the speaker corresponding to the second audio file.

In an embodiment, the voice recognition machine learning may be performed using a voice recognition engine mounted in the set-top box without linking with an external server.

In an embodiment, the step of generating a personalized user interface screen based on the speaker identification result may include determining whether the speaker of the currently activated profile and the identified speaker are the same, and as a result of the determination, if the speaker has changed, It may include configuring the personalized user interface screen corresponding to the profile of the identified speaker based on usage history data previously collected corresponding to the identified speaker.

In an embodiment, the profile may be automatically switched to a profile corresponding to the identified speaker based on a change in the speaker.

In an embodiment, the method further includes extracting a locking keyword by performing phonological analysis and/or voice recognition on the second audio file, and locking control for the corresponding profile is performed based on the extracted locking keyword. It can be.

According to another aspect, storing at least one computer program including instructions that, when executed by at least one processor, cause the at least one processor to perform operations for providing security and personalization services in a set-top box paired with a remote control. In the non-volatile computer-readable storage medium, the operations include receiving a first audio file for each profile from the remote control, registering a security authentication audio file for each profile, and receiving a second audio file from the remote control. Authenticating and identifying a speaker based on the security authentication audio file registered for each profile and the second audio file; generating a personalized user interface screen based on the speaker identification result; and generating a personalized user interface screen. It may include transmitting the screen to the TV.

A set-top box according to another aspect includes a transceiver unit that transmits and receives signals to and from a paired remote control and a TV, and a control unit connected to the transceiver unit, wherein the control unit receives a first audio file from the remote control and provides security authentication audio for each profile. When a file is registered and a second audio file is received from the remote control, the speaker is authenticated and identified based on the security authentication audio file registered for each profile and the second audio file, and the individual is authenticated based on the speaker identification result. A customized user interface screen can be created, and the created personalized user interface screen can be transmitted to the TV.

In an embodiment, the first audio file may include a specific predefined phrase or lock keyword.

In an embodiment, the first audio file and the second audio file may be created in response to a user-selected profile on the home screen of the TV.

In an embodiment, the microphone of the remote control is activated based on the first button provided on the remote control being pressed for a first time while the profile lock setting screen is displayed on the TV, and the microphone of the remote control is activated based on the voice recorded through the activated microphone. Thus, the first audio file can be created.

In an embodiment, while the profile unlock screen is displayed on the TV, the microphone of the remote control is activated based on the first button provided on the remote control being pressed for a second time, and the voice recorded through the activated microphone is activated. Based on this, the second audio file may be created.

In an embodiment, the set-top box further includes an analysis unit that performs phonological analysis on each of the security authentication audio file registered for each profile and the second audio file, and corresponds to the second audio file based on the phonological analysis result. Speakers can be authenticated and identified.

In an embodiment, the analysis unit may perform the phoneme analysis using a phoneme analysis engine mounted on the set-top box without linking with an external server.

In an embodiment, the set-top box extracts a voice feature vector through preprocessing each of the security authentication audio file registered for each profile and the second audio file, and performs voice recognition machine learning based on the extracted voice feature vector. It may further include a learning unit that authenticates and identifies the speaker corresponding to the second audio file.

In an embodiment, the learning unit may perform the voice recognition machine learning using a voice recognition engine mounted in the set-top box without linking with an external server.

In an embodiment, the control unit determines whether the speaker of the currently activated profile and the identified speaker are the same, and when the speaker is changed according to the determination result, based on the usage history data previously collected corresponding to the identified speaker. Thus, the personalized user interface screen corresponding to the profile of the identified speaker can be configured.

In an embodiment, the control unit may automatically switch to a profile corresponding to the identified speaker based on a change in the speaker.

In an embodiment, the control unit may extract a locking keyword based on a result of phonological analysis and/or voice recognition for the second audio file, and perform locking control on the corresponding profile based on the extracted locking keyword.

The technical problems to be achieved in the present invention are not limited to the technical problems mentioned above, and other technical problems not mentioned will be clearly understood by those skilled in the art from the description below. You will be able to.

The present invention has the advantage of providing enhanced security and personalized services using an artificial intelligence-based set-top box.

In addition, the present invention enables speaker identification within the device through the voice recognition function installed inside the set-top box without linking with a separate external server for user identification through voice recognition, thereby providing faster and safer security services and personalized services. There are advantages to this.

In addition, the present invention identifies the speaker through on-device-based voice recognition and automatically unlocks and/or switches profiles, thereby enhancing the security of the personal profile compared to the conventional password input method. In addition, it has the advantage of improving the user experience of personalized services through quick profile lock setting/unlocking and profile switching.

In addition, the present invention has the advantage of further improving the user scenario because it allows a 1-depth profile change through voice recognition, rather than a 2-depth profile change by selecting the conventional home screen entry->profile change menu.

In addition, various effects that can be directly or indirectly identified through this document may be provided.

The drawings attached to this specification are intended to provide an understanding of the present invention, show various embodiments of the present invention, and together with the description of the specification, explain the principles of the present invention.
1 is a diagram illustrating a conventional password-based profile lock setting and unlocking method.
Figure 2 is a system configuration diagram for providing security and personalized services through voice recognition according to an embodiment.
Figure 3 is a flowchart illustrating a method of providing security and personalization services based on on-device voice recognition according to an embodiment.
Figure 4 is a flow chart to explain a method of providing security and personalization services based on voice recognition in a set-top box according to an embodiment.
Figure 5 is a flow chart to explain a method of providing security and personalization services based on voice recognition in a set-top box according to another embodiment.
Figure 6 is a flow chart to explain a method of providing security and personalization services based on voice recognition in a set-top box according to another embodiment.
Figure 7 is a flowchart illustrating a method of providing voice recognition-based security and personalization services through a remote control paired with a set-top box according to an embodiment.
Figure 8 is a diagram for explaining a profile lock setting and unlocking procedure according to an embodiment.
Figure 9 is a block diagram for explaining the structure of a set-top box according to an embodiment.

Hereinafter, some embodiments of the present invention will be described in detail through illustrative drawings. When adding reference numerals to components in each drawing, it should be noted that identical components are given the same reference numerals as much as possible even if they are shown in different drawings. Additionally, when describing embodiments of the present invention, if detailed descriptions of related known configurations or functions are judged to impede understanding of the embodiments of the present invention, the detailed descriptions will be omitted.

In describing the components of the embodiments of the present invention, terms such as first, second, A, B, (a), and (b) may be used. These terms are only used to distinguish the component from other components, and the nature, sequence, or order of the component is not limited by the term. Additionally, unless otherwise defined, all terms used herein, including technical or scientific terms, have the same meaning as generally understood by a person of ordinary skill in the technical field to which the present invention pertains. Terms defined in commonly used dictionaries should be interpreted as having a meaning consistent with the meaning in the context of the related technology, and should not be interpreted in an ideal or excessively formal sense unless explicitly defined in the present application. No.

In various examples of this disclosure, “/” and “,” should be interpreted as indicating “and/or.” For example, “A/B” can mean “A and/or B.” Furthermore, “A, B” may mean “A and/or B.” Furthermore, “A/B/C” may mean “at least one of A, B and/or C.” Furthermore, “A, B, C” may mean “at least one of A, B and/or C.”

In various examples of this disclosure, “or” should be interpreted as indicating “and/or.” For example, “A or B” may include “only A,” “only B,” and/or “both A and B.” In other words, “or” should be interpreted as indicating “additionally or alternatively.”

Hereinafter, embodiments of the present invention will be described in detail with reference to FIGS. 1 to 9.

1 is a diagram illustrating a conventional password-based profile lock setting and unlocking method.

Referring to FIG. 1, the user can select his/her profile image among profile images displayed on the TV screen (S110).

The user can set the lock setting password on the profile lock password setting screen (S120). At this time, the user can enter the password using the number button provided on the remote control.

With multiple profiles registered, the user can select the profile image he or she wants to unlock among the locked profile images displayed on the home screen (S130).

The user can unlock the profile by entering the preset lock password corresponding to the profile image selected on the profile unlock screen (S140). At this time, the user can enter the password using the number button provided on the remote control.

If the user does not remember the password and wishes to re-register the password corresponding to his or her profile, the user can use the delete button 131 to delete the profile and then reset the profile lock password through the re-registration procedure.

As described above, the conventional set-top box profile unlocking procedure provides a two-depth unlocking procedure of entering the home screen and entering the lock password on the profile unlock screen, which not only reduces user convenience but also reduces profile lock settings. Due to the change, there was a problem that interfered with TV viewing.

Figure 2 is a system configuration diagram for providing security and personalized services through voice recognition according to an embodiment.

Referring to FIG. 2, the system 1 includes a TV 10, a set-top box 20, a remote control 30, a digital broadcasting server 40, a voice recognition server 50, and a communication network 60. You can.

The set-top box 20 can be connected to the TV 10 through a dedicated line - for example, an HDMI cable - and can separate and decode the signal received from the digital broadcasting server 40 and provide it to the TV 10. .

As an example, the remote control 30 may be paired with the set-top box 20 and the TV 10 at the same time and may be equipped with an integrated remote control function that can transmit a wireless control signal to the set-top box 20 or the TV 10.

As another example, the remote control 30 is paired with the set-top box 20, and a remote control (not shown) to control the TV 10 may be provided separately.

The remote control 30 has a power button for controlling the power ON/OFF of the set-top box 20 and the TV 10, and a user voice input for setting/unlocking/switching the profile lock and transmitting it to the set-top box 20. It may include, but is not limited to, a push-to-talk button, a channel selection button, a voice control button, a search button, a display adjustment button, an external input control button, a home menu selection button, and a settings button. No.

The user can use the settings button provided on the remote control 30 to enter the settings menu on the home screen, and select specific items in the settings menu to perform profile registration/deletion procedures, user-specific lock keyword registration procedures, and voice recognition-based profiles. You can perform lock setting/unlocking procedures, locking keyword registration procedures for each speaker, and voice recognition-based profile automatic switching setting procedures.

The set-top box 20 may be equipped with a microphone and a speaker, and when the voice recognition function of the set-top box 20 is activated, the user voice information input through the microphone is transmitted to the voice recognition server 50 through the communication network 60. can be sent to The voice recognition server 50 may perform voice recognition based on the received user voice information and then control the result corresponding to the recognized user command to be transmitted to the set-top box 20 through the communication network 60. The set-top box 20 can output the received results through a provided speaker.

In an embodiment, when a specific button provided on the remote control 30 - for example, a Push-to-Talk button - is pressed by the user while the microphone of the set-top box 20 is activated for voice recognition, the remote control 30 ) can output a predetermined control signal to deactivate the microphone of the set-top box 20. User voice commands input through the microphone provided in the remote control 30 - for example, voice keywords for profile lock setting, profile unlock, and automatic profile switching - may be transmitted to the set-top box 20.

As an example, voice keywords for profile lock setting, profile unlock, and profile automatic switching may be implemented to be the same for all users, but this is only one embodiment, and different keywords will be applied for each user depending on user settings. It may be possible.

For example, the voice keyword for setting the profile lock can be predefined as “Set the profile lock,” the voice keyword for unlocking the profile can be predefined as “Unlock the profile,” and the voice keyword for switching profiles can be predefined as “Switch profile.” However, this is only one embodiment, and each user may predefine and use voice keywords corresponding to each operation.

In an embodiment, while the user's profile is locked, the user can unlock the profile by uttering a predefined lock keyword while pressing a specific button on the remote control 30. Here, when a specific button for entering a lock keyword - for example, a erase button - is pressed, the microphone of the remote control 30 is activated, and the microphone of the set-top box 20 can be deactivated while the specific button is pressed. . In an embodiment, the remote control 30 button for entering the lock keyword may be predefined as a different button according to the design of a person skilled in the art, or may be changed through a predetermined setting menu.

In another embodiment, while the user's profile is locked, the user selects the profile change menu on the home screen and then presses a specific button on the remote control 30 - for example, the erase button - and utters a predefined lock keyword to You can unlock the profile. Here, when a specific button for inputting a lock keyword is pressed, the microphone of the remote control 30 may be activated, and the microphone of the set-top box 20 may be deactivated while the specific button is pressed.

In another embodiment, while the user's profile is locked, the user may select the profile change menu on the home screen and then utter the lock keyword without pressing a specific button on the remote control 30 to unlock the profile. there is. In this case, the voice uttered by the user can be directly input to the microphone of the set-top box 20 and immediately detected, and if the input user voice is a preset or pre-registered lock keyword, the set-top box 20 You can unlock the user's profile.

In an embodiment, a voice recording file for each user for a lock keyword may be pre-registered and maintained in a predetermined recording area of the set-top box 20. In a state in which the lock setting for a plurality of profiles is released, the set-top box 20 can recognize the lock keyword uttered by the user by voice. The set-top box 20 can identify the speaker who uttered the lock keyword by analyzing the pre-registered voice recording file for each user corresponding to the recognized lock keyword. The set-top box 20 can configure a customer-customized screen corresponding to the profile of the identified speaker. In an embodiment, a customer-customized screen may be dynamically configured based on service usage data and statistics collected in advance for each user. For example, the customer-customized screen may include content continuation information, recommended content curation images, information on recommended genres, etc., but is not limited to this.

Through the above-described system 1, the present invention allows the user to press a specific button on the remote control 30, or enter a lock keyword after pressing a specific button, without entering the home screen when a profile change is necessary even while watching TV. You have the advantage of being able to receive personalized services by uttering a message and immediately changing it to your profile.

In addition, the present invention has the advantage of further improving the user scenario because it allows a 1-depth profile change through voice recognition, rather than a 2-depth profile change by selecting the conventional home screen entry->profile change menu.

Figure 3 is a flowchart illustrating a method of providing security and personalization services based on on-device voice recognition according to an embodiment.

Referring to FIG. 3, when setting a profile lock, the remote control 30 can create a security authentication audio file for security authentication of the profile and transmit it to the set-top box 20 (S301). As an example, a security authentication audio file may be created as a PCM (Pulse Code Modulation) type. Here, the security authentication audio file may be a file in which security authentication keywords are repeatedly recorded a certain number of times to improve security authentication accuracy.

When the profile lock control button is selected by the user, the remote control 30 activates the microphone provided on the remote control 30 to start voice recording, and when the profile lock control button is deselected, the remote control 30 creates a recorded audio file. (S303 to S309). In an embodiment, when the profile lock control button is selected by the user, the remote control 30 transmits a predetermined control signal - for example, a security authentication request signal - to the set-top box 20, and the set-top box 20 transmits the received A UI screen that guides utterance of a specific lock keyword according to a control signal can be created and output on the screen of the connected TV 10. For example, the locking keyword may be “high LG,” but it is not limited thereto, and the locking keyword may be changed according to the design or user settings of a person skilled in the art.

The remote control 30 can transmit the recorded audio file to the set-top box 20 (S309).

The set-top box 20 can extract keywords by performing voice recognition based on the received audio file (S311).

The set-top box 20 may determine whether the extracted keyword is a predefined locking keyword (S313).

As a result of the determination, if the extracted keyword is a lock keyword, the set-top box 20 runs the installed speaker authentication and identification app to perform speaker authentication and identification operations based on the pre-registered security authentication audio file and the received audio file. (S315).

As an example, speaker authentication and identification operations may be performed through machine learning based on a predefined learning model. That is, speaker authentication and identification can be achieved through comparative analysis of the voice feature learning results for each speaker for pre-registered security authentication audio files and the speaker's voice feature learning results corresponding to the received audio files. For example, voice features to be learned may include, but are not limited to, gender, age, frequency, amplitude, waveform, speech rate, time difference between syllables, etc. For example, the set-top box 20 may perform received signal amplification and noise filtering operations to improve the sound quality of the received audio file, and may extract a voice feature vector for the filtered audio signal. Thereafter, the set-top box 20 can recognize keywords through recurrent neural network learning on the extracted voice feature vector, and authenticate and identify the speaker through deep neural network learning on the extracted voice feature vector.

If speaker authentication and identification are successful, the set-top box 20 may configure a personalized UI screen for the identified speaker based on service usage data previously collected corresponding to the identified speaker (S317).

The configured screen may be transmitted from the set-top box 20 to the TV 10 and output on the screen (S319 to S321).

As a result of the determination in step 313, if the extracted keyword does not match the preset lock keyword, the set-top box 20 may output a predetermined voice recognition failure message (S323). For example, a voice recognition failure message may be output through the speaker of the set-top box 20, but this is only one embodiment, and in another embodiment, it may be output through the screen of the TV 10.

Figure 4 is a flow chart to explain a method of providing security and personalization services based on voice recognition in a set-top box according to an embodiment.

Referring to FIG. 4, the set-top box 20 may register at least one profile in the internal database according to the user menu selection on the home screen of the TV 10 (S410).

When the security authentication audio file created corresponding to the profile registered during the initial profile lock setting is received from the remote control, the set-top box 20 registers the security authentication audio file corresponding to the profile (or user or speaker) in the internal database. You can do it (S420).

The set-top box 20 can receive an audio file recorded with the microphone of the remote control 30 from the remote control 30 (S430). As an example, when a specific profile lock control button is selected, the remote control 30 activates the provided microphone to start voice recording, and when the profile lock control button is deselected, the remote control 30 converts the recorded voice into an audio file and stores it in the set-top box (20). ) can be transmitted to.

The set-top box 20 can authenticate and identify the speaker by comparing the phonological analysis result for the received audio file with the phonological analysis result for the security authentication audio file for each registered speaker (S440).

The set-top box 20 may configure a personalized UI screen corresponding to the identified speaker based on the difference between the identified speaker and the speaker corresponding to the currently activated profile (S450).

The set-top box 20 can transmit the configured screen to the TV (S460).

As described above, in the present invention, the set-top box 20 authenticates and identifies the speaker through voice recognition for the user's voice input through the remote control microphone, and then displays a personalized UI screen corresponding to the identified speaker on the TV 10. There is an advantage in that the user does not have to change the speaker or control the profile lock by selecting a separate menu on the home screen of the TV 10.

Figure 5 is a flow chart to explain a method of providing security and personalization services based on voice recognition in a set-top box according to another embodiment.

Referring to FIG. 5, the set-top box 20 can register at least one profile according to a user menu selection on the home screen of the TV 10 (S510).

When setting the initial profile lock, the set-top box 20 may receive a lock keyword audio file created corresponding to the profile selected by the user from the remote control 30 and register it in the internal database (S520).

The set-top box 20 can extract the lock keyword by recognizing the audio file received from the remote controller 30 according to the user's selection of the profile lock setting control button (S530).

The set-top box 20 can authenticate and identify the speaker corresponding to the received audio file by comparing the phonological analysis result for the received audio file with the phonological analysis result for the pre-registered lock keyword audio file (S540).

The set-top box 20 may determine whether the speaker of the currently activated profile and the identified speaker are the same (S550).

If the determination result in step 550 is not the same, the set-top box 20 may configure a personalized UI screen based on previously collected usage history data corresponding to the identified speaker (S560).

The set-top box 20 can convert the currently active profile to the profile of the identified speaker and then transmit the configured personalized UI screen to the TV 10 (S570).

As a result of the determination in step 550, if the speaker of the currently activated profile is the same as the identified speaker, the set-top box 20 may control profile lock settings corresponding to the identified speaker based on the extracted lock keyword (S580) ). For example, if the extracted lock keyword is 'profile lock setting', the set-top box 20 can lock and deactivate the currently activated profile. As another example, if the extracted lock keyword is 'unregister profile', all registration information for the currently activated profile can be deleted. As another example, if the extracted lock keyword is 'security authentication release', the set-top box 20 may delete the lock keyword audio file already registered for the currently activated profile from the database.

Figure 6 is a flow chart to explain a method of providing security and personalization services based on voice recognition in a set-top box according to another embodiment.

Referring to FIG. 6, when the set-top box 20 sets a lock for a specific profile by selecting a predetermined menu on the TV home screen, a security authentication audio file containing a specific predefined phrase is sent from the remote control (or to the set-top box). It can be received (through the provided microphone) and registered in the internal database (S610). As an example, a specific phrase may be “one, two, three, four, five, six, seven, eight, nine, ten,” but is not limited thereto and may be changed according to the design or user settings of a person skilled in the art.

The set-top box 20 provides a voice recognition UI that includes a predetermined notification message requesting the user to utter a security authentication keyword based on a specific button on the remote control 30 - for example, the erase button - being pressed for a certain period of time while watching TV. It can be generated and displayed on one side of the TV 10 screen (S620).

Based on the audio file recorded through the microphone of the remote control 30 according to the notification message received from the remote control 30, the set-top box 20 provides phonology for the received audio file and the security authentication audio file for each profile registered in the database. Analysis can be performed (S630).

The set-top box 20 can authenticate and identify the speaker corresponding to the received audio file based on the phoneme analysis result and then determine whether the speaker has changed (S640).

The set-top box 20 may deactivate the existing speaker's profile based on a change in the speaker and automatically activate the changed speaker's profile (S650). Here, deactivation of the profile may mean that the profile is automatically converted from the unlocked state to the locked state, and activation of the profile may mean that the profile is automatically converted from the locked state to the unlocked state.

The set-top box 20 can configure a personalized UI screen corresponding to the automatically activated profile and display it on the TV 10 screen (S660).

Figure 7 is a flowchart illustrating a method of providing voice recognition-based security and personalization services through a remote control paired with a set-top box according to an embodiment.

Referring to FIG. 7, when a first button provided on the remote control 30 - for example, an OK button - is pressed by the user for a first time, the remote control 30 activates a microphone provided to perform a predefined security operation. The user's voice for the authentication phrase (or lock keyword) can be recorded (S710). Here, the user's voice for the security authentication phrase may be repeatedly recorded a preset number of times. When the first button is pressed for a long time for a first time, the first button may be used as a profile lock setting button rather than as a general confirmation button.

When recording is completed, the remote control 30 may create a first audio file based on the recorded voice (S720). Here, the first audio file may be created as a PCM type.

The remote control 30 can transmit the generated first audio file to the set-top box 20 (S730). The set-top box 20 may register the first audio file as a security authentication audio file corresponding to the profile and store it in an internal database.

In an embodiment, security authentication audio files for multiple profiles may be registered and maintained in the set-top box 20 through the remote control 30.

When the second button provided on the remote control 30 - for example, the erase button - is pressed by the user for a second time, the remote control 30 activates the microphone provided and produces second audio corresponding to the user's input voice. A file can be created (S740). Here, the second audio file may be created in PCM type. When the second button is pressed and held for a second time, the second button may be used as a profile lock control button rather than as a general erase button.

The remote control 30 can transmit the generated second audio file to the set-top box 20 (S750).

The set-top box 20 according to the embodiment may perform speaker authentication and identification operations through phoneme analysis of the first audio file registered for each profile and the received second audio file.

In another embodiment, the set-top box 20 may extract a voice feature vector through preprocessing of the first audio file registered for each profile and the received second audio file. The set-top box 20 can recognize keywords through recurrent neural network learning on the extracted voice feature vector, and authenticate and identify the speaker through deep neural network learning on the extracted voice feature vector.

The set-top box 20 can determine whether the speaker has changed, configure a personalized UI screen corresponding to the changed speaker based on the speaker change, and display it on the TV 10 screen.

Figure 8 is a diagram for explaining a profile lock setting and unlocking procedure according to an embodiment.

Referring to FIG. 8, the user can select his/her profile image among profile images displayed on the TV screen (S810). At this time, a profile lock setting screen may be displayed on the TV screen.

The user can utter a keyword for setting the profile lock according to the guidance message on the profile lock setting screen (S820). At this time, the keyword uttered by the user may be recorded through a microphone provided in the remote control 30 and then transmitted to the set-top box 20 and registered as a security authentication audio file for the user (or profile).

While watching TV, the user can select one of the locked profiles (S830). At this time, a profile unlock screen may be displayed on the TV screen.

The user may utter a keyword for unlocking while pressing a specific button - for example, a confirmation button or a clear button - according to the guidance message on the profile lock setting screen (S840). At this time, the keyword uttered by the user may be recorded through a microphone provided in the remote control 30 and then transmitted to the set-top box 20. The set-top box 20 can authenticate and identify the speaker by performing phonological analysis and/or voice recognition machine learning on the pre-registered security authentication audio file corresponding to the profile and the audio file recorded on the profile lock setting screen. .

Figure 9 is a block diagram for explaining the structure of a set-top box according to an embodiment.

Referring to FIG. 9, the set-top box 20 is configured to include a transceiver 910, a learning unit 920, a control unit 950, an analysis unit 930, a storage unit 940, and a control unit 950. You can.

The transmitting and receiving unit 910 includes a first communication unit 911 for transmitting/receiving signals with the remote control 30, a second communication unit 912 for transmitting/receiving signals with the TV 10, and a signal transmission unit 910 with the server 910. /It may be configured to include a third communication unit 913 for reception.

The control unit 950 can control the overall operation and input/output of the set-top box 20.

The control unit 950 may receive an audio file transmitted by the remote control 30 from the transceiver 910.

If the received audio file is a security authentication audio file, the control unit 950 may register it in the storage unit 940 as a security authentication audio file corresponding to the corresponding profile.

If the received audio file is an audio file for unlocking, the control unit 950 may provide the corresponding audio file to the learning unit 920 and/or the analysis unit 930.

The learning unit 920 can perform voice recognition machine learning based on the audio file received from the control unit 950 and the security authentication audio file previously registered for each profile, and perform speaker authentication and speaker identification based on the learning results. there is.

The control unit 920 can determine whether to change the speaker based on the speaker authentication and identification results of the learning unit 920, and can configure a personalized UI screen based on usage history data for each profile according to the judgment result. At this time, the configured personalized UI screen may be transmitted and displayed on the TV 10 through the second communication module 912 of the transceiver 910.

The analysis unit 930 may perform phonological analysis based on the audio file received from the control unit 950 and the security authentication audio file previously registered for each profile. The analysis unit 930 may perform speaker authentication and speaker identification based on the phonological analysis results.

The control unit 920 can determine whether to change the speaker based on the speaker authentication and identification results of the analysis unit 930, and can configure a personalized UI screen based on usage history data for each profile according to the judgment result. At this time, the configured personalized UI screen may be transmitted and displayed on the TV 10 through the second communication module 912 of the transceiver 910.

The control unit 920 may generate usage history data and/or usage history statistical data for each profile collected while watching TV and store them in the storage unit 940 .

Based on the usage history data and/or usage history statistical data for each profile, the control unit 920 creates a list of channels and content to be continued, a recommended genre VOD list, a recommended broadcast channel list, a recommended YouTube channel list, etc. corresponding to the profile. You can create a personalized UI screen by creating information.

When an audio file for profile lock control is received from the remote controller 30, the control unit 920 does not request voice recognition for the audio file from the external voice recognition server 50, but uses the internally mounted voice recognition engine and/or Alternatively, secure authentication and identification procedures for the speaker can be performed using the phonological analysis engine.

The control unit 920 can determine whether a speaker has changed through speaker authentication and identification based on the speaker's voice input through the remote control 30, and when a speaker change is confirmed, the profile is automatically switched to create a new speaker corresponding to the current speaker. You can create custom UI screens.

The storage unit 940 may be equipped with a software engine for voice recognition and/or phonological analysis - that is, an app.

The storage unit 940 may maintain usage history data and security authentication audio files for each profile in the form of a database.

The learning unit 920 and/or the analysis unit 930 may identify the locking keyword uttered by the user based on the audio file received from the remote control 30.

The control unit 950 may control lock settings for the profile based on the identified lock keyword.

Additionally, the control unit 950 may delete the security authentication audio file and/or usage history data registered for the profile based on the identified lock keyword.

The server described in relation to the embodiments disclosed in this specification is connected to a display, a transceiver for transmitting and receiving signals with network equipment and user devices - for example, a set-top box - through an external wired / wireless communication network, and the display and transceiver are connected to the overall It may be configured to include at least one processor that controls the operation and a memory in which a program for the operation of the processor is recorded.

The set-top box described in relation to the embodiments disclosed herein can be connected to an external device - for example, a digital broadcasting server, a voice recognition server, It may be configured to include a transceiver for transmitting and receiving signals to and from a remote control, TV, etc., at least one processor connected to the transceiver to control the overall operation, and a memory in which a program for the operation of the processor is recorded.

The steps of the method or algorithm described in connection with the embodiments disclosed herein may be implemented directly as hardware, software modules, or a combination of the two executed by a processor. Software modules may reside in a storage medium (i.e., memory and/or storage) such as RAM memory, flash memory, ROM memory, EPROM memory, EEPROM memory, registers, hard disk, removable disk, CD-ROM.

An exemplary storage medium is coupled to a processor, the processor capable of reading information from and writing information to the storage medium. Alternatively, the storage medium may be integral with the processor. The processor and storage medium may reside within an application specific integrated circuit (ASIC). The ASIC may reside within the user device. Alternatively, the processor and storage medium may reside as separate components within the user device.

The above description is merely an illustrative explanation of the technical idea of the present invention, and various modifications and variations will be possible to those skilled in the art without departing from the essential characteristics of the present invention.

Accordingly, the embodiments disclosed in the present invention are not intended to limit the technical idea of the present invention, but rather to explain it, and the scope of the technical idea of the present invention is not limited by these embodiments. The scope of protection of the present invention should be interpreted in accordance with the claims below, and all technical ideas within the equivalent scope should be construed as being included in the scope of rights of the present invention.

Claims (25)

In a method of providing security and personalization services in a set-top box paired with a remote control,
Receiving a first audio file for each profile from the remote control and registering a security authentication audio file for each profile;
Receiving a second audio file from the remote control;
authenticating and identifying a speaker based on the security authentication audio file registered for each profile and the second audio file;
generating a personalized user interface screen based on the speaker identification result; and
Transmitting the generated personalized user interface screen to a TV;
Including,
The first audio file is created by repeatedly recording a predefined locking keyword a preset number of times by the speaker,
The lock keyword is defined identically for all profiles in the set-top box,
method. delete According to paragraph 1,
The method wherein the first audio file and the second audio file are created corresponding to a user-selected profile on the home screen of the TV. According to paragraph 1,
The microphone of the remote control is activated based on the first button provided on the remote control being pressed for a first time while the profile lock setting screen is displayed on the TV, and the microphone of the remote control is activated based on the voice recorded through the activated microphone. 1 How audio files are created. According to paragraph 1,
With the profile unlock screen displayed on the TV, the microphone of the remote control is activated based on the first button provided on the remote control being pressed for a second time, and the microphone of the remote control is activated based on the voice recorded through the activated microphone. A method by which a second audio file is created. According to paragraph 1,
The step of authenticating and identifying the speaker based on the security authentication audio file registered for each profile and the second audio file,
performing phonological analysis on each of the security authentication audio file registered for each profile and the second audio file; and
Authenticating and identifying the speaker corresponding to the second audio file based on the phonological analysis results
Method, including. According to clause 6,
The method is characterized in that the phonological analysis is performed using a phonological analysis engine mounted on the set-top box without linking with an external server. According to paragraph 1,
The step of authenticating and identifying the speaker based on the security authentication audio file registered for each profile and the second audio file,
Extracting a voice feature vector through preprocessing each of the security authentication audio file registered for each profile and the second audio file; and
Authenticating and identifying the speaker corresponding to the second audio file by performing voice recognition machine learning based on the extracted voice feature vector.
Method, including. According to clause 8,
A method characterized in that the voice recognition machine learning is performed using a voice recognition engine mounted in the set-top box without linking with an external server. According to paragraph 1,
The step of generating a personalized user interface screen based on the speaker identification result includes:
determining whether the speaker of the currently activated profile and the identified speaker are the same; and
As a result of the determination, if the speaker has changed, configuring the personalized user interface screen corresponding to the profile of the identified speaker based on usage history data previously collected corresponding to the identified speaker.
Method, including. According to clause 10,
A method, characterized in that automatically switching to a profile corresponding to the identified speaker based on a change in the speaker. According to paragraph 1,
Further comprising extracting a second locking keyword by performing phonological analysis and/or voice recognition on the second audio file, wherein locking control for the corresponding profile is performed based on the second locking keyword. How to. A non-volatile computer storing at least one computer program containing instructions that, when executed by at least one processor, cause the at least one processor to perform operations for providing security and personalization services in a set-top box paired with a remote control. In a readable storage medium,
The above operations are:
Receiving a first audio file for each profile from the remote control and registering a security authentication audio file for each profile;
Receiving a second audio file from the remote control;
authenticating and identifying a speaker based on the security authentication audio file registered for each profile and the second audio file;
generating a personalized user interface screen based on the speaker identification result; and
Transmitting the generated personalized user interface screen to a TV;
Including,
The first audio file is created by repeatedly recording a predefined locking keyword a preset number of times by the speaker,
The lock keyword is defined identically for all profiles in the set-top box,
storage media. In a set-top box to provide security and personalized services,
A transceiver unit that transmits and receives signals with the paired remote control and TV; and
It includes a control unit connected to the transceiver unit,
The control unit receives a first audio file from the remote control and registers a security authentication audio file for each profile, and when a second audio file is received from the remote control, the security authentication audio file and the second audio file registered for each profile. Authenticate and identify the speaker based on, generate a personalized user interface screen based on the speaker identification result, and transmit the generated personalized user interface screen to the TV,
The first audio file is created by repeatedly recording a predefined locking keyword a preset number of times by the speaker,
The lock keyword is defined identically for all profiles in the set-top box,
Set top box. delete According to clause 14,
A set-top box, characterized in that the first audio file and the second audio file are created in response to a user-selected profile on the home screen of the TV. According to clause 14,
The microphone of the remote control is activated based on the first button provided on the remote control being pressed for a first time while the profile lock setting screen is displayed on the TV, and the microphone of the remote control is activated based on the voice recorded through the activated microphone. 1 A set-top box, characterized in that audio files are created. According to clause 14,
With the profile unlock screen displayed on the TV, the microphone of the remote control is activated based on the first button provided on the remote control being pressed for a second time, and the microphone of the remote control is activated based on the voice recorded through the activated microphone. A set-top box, characterized in that a second audio file is created. According to clause 14,
Further comprising an analysis unit that performs phonological analysis on each of the security authentication audio file registered for each profile and the second audio file, wherein the speaker corresponding to the second audio file is authenticated and identified based on the phonological analysis result. A set-top box characterized by: According to clause 19,
A set-top box, characterized in that the analysis unit performs the phoneme analysis using a phoneme analysis engine mounted on the set-top box without linking with an external server. According to clause 14,
A voice feature vector is extracted through preprocessing for each of the security authentication audio file registered for each profile and the second audio file, and voice recognition machine learning is performed based on the extracted voice feature vector to correspond to the second audio file. A set-top box further comprising a learning unit that authenticates and identifies the speaker. According to clause 21,
A set-top box, characterized in that the learning unit performs the voice recognition machine learning using a voice recognition engine mounted in the set-top box without linking with an external server. According to clause 14,
The control unit determines whether the speaker of the currently activated profile and the identified speaker are the same, and when the speaker changes according to the determination result, the control unit identifies the speaker based on usage history data previously collected corresponding to the identified speaker. A set-top box, characterized in that configuring the personalized user interface screen corresponding to the profile of the speaker. According to clause 23,
The set-top box is characterized in that the control unit automatically switches to a profile corresponding to the identified speaker based on a change in the speaker. According to clause 14,
The control unit extracts a second lock keyword based on a result of phoneme analysis and/or voice recognition for the second audio file, and performs lock control for the corresponding profile based on the second lock keyword. , set-top box.