KR20010056900A - A digital television receiver having the speaker recognition module and a speaker recognition method thereof - Google Patents

Abstract

PURPOSE: A digital TV receiver equipped with a speaker recognition module and a method for recognizing a speaker thereof are provided to increase the convenience of a user by enabling a user admitted using a speaker recognition to selectively watch television. CONSTITUTION: A tuner(101) receives a signal of a desired channel from an antenna. A packet separator(102) respectively separates transport signal lines into a video signal line, an audio signal line and an addition information line. An audio decoder and a video decoder(103,104) extend the audio and the video signal line separated. A microcomputer(108) receives the additional information line. A user interface(112) transmits a user's command. An OSD(On Screen Display) processor(107) receives OSD information relevant to a key input from the user interface(112). A switch(105) switches outputs of the video decoder(104) and the OSD processor(107). An NTSC(National Television Standards Committee) encoder(106) outputs a video signal transmitted through the switch(105). A speech input unit(113) inputs the user's speech. A speaker recognition module(114) extracts features of the user's speech inputted through the speech input unit(113).

Description

A digital TV receiver having a speaker recognition module and a speaker recognition method thereof {A DIGITAL TELEVISION RECEIVER HAVING THE SPEAKER RECOGNITION MODULE AND A SPEAKER RECOGNITION METHOD THEREOF}

The present invention relates to an apparatus and method for speaker recognition of a digital TV receiver, which enables an operator to recognize a user's identity on a digital TV. In particular, the present invention can selectively access a TV only to an authorized user using speaker recognition. The present invention relates to a speaker recognition apparatus and method of a digital TV receiver.

Currently, digital TV broadcasting includes digital satellite broadcasting through satellites and terrestrial digital broadcasting using terrestrial waves, and cable digital broadcasting will also be implemented.

In these three broadcasting schemes, the transmission media (satellite, terrestrial, cable) are distinctly different, but the remainder is almost the same, including video / audio compression and packet composition.

Therefore, the configuration of these broadcast receivers also differs only in the parts related to the transmission scheme, and the remaining parts are almost similar.

Currently, the basic transmission format of digital broadcasting is MPEG-2, which is defined as an international standard.It is based on ISO / IEC 13818-1 (system), ISO / IEC 13818-2 (video), and ISO / IEC 13818-3 (audio) standards. It is prescribed.

However, US and Korean digital TV receivers do not use MPEG audio in the audio standard, but follow the Dolby AC-3 audio standard.

The system specification defines a time-multiplexed signal sequence called Transport Stream Packet (TSP), and the video specification and audio specification define a compressed video and audio signal sequence.

The header at the beginning of the packet contains a PID (Packet IDdentifier) number.

The PID number provides information for demultiplexing a time multiplexed signal.

The PID number is an identification number indicating whether the current packet is a video packet, an audio packet, or additional information (Program Specific Information).

Video, audio and other additional information are time-multiplexed and carried in several transport stream packets, and these signal sequences are distinguished by PID numbers.

As described above, when multiplexed transport stream packets are transmitted to a digital TV receiver, the digital TV receiver tunes a desired channel to restore the original video / audio signal so that the viewer can watch the video displayed on the screen. The voice corresponding to the image output to the speaker is heard.

A digital TV receiver operating as described above will be described with reference to FIG. 1.

1 is a block diagram of a general digital TV receiver. As shown therein, a tuner 1 for receiving a signal of a desired channel from an antenna, and outputting a transport signal sequence for detecting and error-correcting the signal; A packet separator 2 for receiving the transport signal sequence outputted from the video signal sequence, an audio signal sequence, and an additional information signal sequence, and an audio decoder for extending and outputting the audio signal sequence output from the packet separator 2; ), A video decoder 4 which expands and outputs a video signal sequence output from the packet separator 2, and converts a digital audio signal output from the audio decoder 3 into an analog audio signal to convert the L signal of the speaker. , Audio-digital-to-analog converter 11 to be transmitted to R, a microcomputer 8 for controlling the whole system, and a program of the microcomputer 8 A ROM (9) storing a gram, a RAM (10) used as a storage location for additional information necessary for the execution process, a user interface (12) for transmitting a user's desired command to the microcomputer (8), and the user interface An OSD processor (7) for receiving OSD information corresponding to a key input from the ROM (12) and outputting an OSD image, and a switch for switching the outputs of the video decoder (4) and the OSD processor (7); 5) and an NTC encoder 6 which outputs a luminance signal Y and a color signal C so that the video signal transmitted through the switch 5 can be displayed on a general TV or monitor CRT. do.

Looking at the prior art configured in this way in detail as follows.

The tuner 1 receives a VSB (Vestigial SideBand) or QAM (Quadrature Amplitude Modulation) signal from an antenna, detects the signal, and outputs a transport signal sequence obtained by correcting an error to the packet separator 2.

The packet separator 2 receives the transport signal sequence and demultiplexes the video signal sequence, the audio signal sequence, and the additional information signal sequence.

The audio signal sequence of the separated signal sequence is provided to the audio decoder 3, the video signal sequence is provided to the video decoder 4, and the additional information signal sequence is provided to the microcomputer 8.

Then, the audio decoder 3 expands the voice signal sequence compressed with Dolby AC3 to the original signal and outputs the audio signal to the audio digital / analog converter 11. The audio digital / analog converter 11 converts the digital voice signal into an analog voice. It converts the signal into L and R of the speaker and outputs the sound.

The video decoder 4 then decompresses the MPEG-2 compressed video signal string to its original size and outputs it to the NTC encoder 6.

At this time, the microcomputer 8 reads the additional information separated from the packet separator 2, analyzes and combines it into information that can be provided to the viewer, and stores the data in the RAM 10 for use as needed. .

When the function key is input through the user interface 12 such as a remote controller and an operation key, the microcomputer 8 performs an operation for performing the function.

The microcomputer 8 reads the OSD information corresponding to the key input from the user interface 12 into the ROM 9 and transmits it to the OSD processor 7.

The OSD processor 7 processes the OSD information into an OSD image and outputs the OSD information to the switch 5.

At this time, the microcomputer 8 controls the switch 5 to switch between the video decoder 4 and the OSD processor 7 alternately.

The video decoder (4) and the OSD processor (7) receive an image and OSD as input from the NTC encoder (6) to process the luminance signal (Y) and color signal (C) to display on a general TV or monitor Output

This output signal is input to the CRT to show an image.

Among the additional information for the US and Korean digital TVs, the PSIP (Program and System Information Protocol) is included.

The PSIP consists of a combination of the following tables, the basic structure of which is shown in FIG.

MGT (Master Guide Table): Version control of all tables

TVRES (Terrestrial Virtual Channel Table): Stores information about channels

RRT (Rating Region Table): Stores regional rating tables for each program

Event Information Table (EIT): Information about current and future programs

Extended Text Table (ETT): Details of current and future programs

STT (System Time Table): Send current time

The broadcasting station indicates the definition of the program rating for each region on the RRT, and sends the program rating information in the content advisory descriptor of the EIT.

In digital broadcast reception, the content advisory descriptor is decoded to determine whether to allow a user to access a program.

As a technology related to the rating of the viewer, it is possible to specifically set the information on the rating by setting the lock menu, and access the lock menu using a password input method through the user interface 12. .

Here, the rock menu is a kind of program that should not be viewed by minors but only by adults.

However, in the prior art as described above, when a viewer wants to access a lock menu, a password input method is used. There is a risk of password leakage, and there is inconvenience in using a password. If you forget your password, you will not be able to use it until you find it.

Accordingly, an object of the present invention for solving the conventional problems as described above is to digitally implement the contents of the RRT (Rating Region Table) included in the PSIP (Program and System Information Protocol), which is additional information of the digital TV. The present invention provides a speaker recognition apparatus and method for a TV receiver.

Another object of the present invention is to provide an apparatus and method for speaker recognition of a digital TV receiver capable of recognizing an operator's identity on a digital TV without the inconvenience of the operator.

It is still another object of the present invention to provide a speaker recognition apparatus and method for a digital TV receiver, which enables a user to selectively watch TV only by an authorized user using speaker recognition.

1 is a block diagram of a general digital TV receiver.

2 is a basic structural diagram of a program and system information protocol (PSIP) table in FIG. 1;

Figure 3 is a block diagram of a digital TV receiver having a speaker recognition module of the present invention.

4 is a speaker recognition process by the speaker recognition module in FIG.

5 is a flowchart illustrating a method for registering a user upon initial shipment of a digital TV receiver having a speaker recognition module of the present invention.

FIG. 6 is a flowchart illustrating a method for extracting user features when registering a user in FIG. 5.

Figure 7 is a first embodiment of the speaker recognition method of the present invention digital TV receiver.

8 is a second embodiment of a speaker recognition method of a digital TV receiver.

***** Explanation of symbols for the main parts of the drawing *****

101: tuner 102: packet separator

103: Audio Decoder 104: Video Decoder

105: switch 106: NTS encoder

107: OSD Processor 108: Microcomputer

109: Rom 110: Ram

111: Audio Digital / Analog Converter

112: user interface 113: voice input unit

114: speaker recognition module

In order to achieve the above object, the present invention provides a first step of muting audio when a user accesses a lock menu, a second step of accepting a user voice after audio muting, and receiving audio after receiving a user voice. A third step of searching whether the user is a registered speaker through the speaker recognition module after muting; and a fourth step of allowing the access to the lock menu if the speaker is registered above and not to access the lock menu unless the speaker is registered. Characterized in consisting of steps.

In addition, the present invention is a first step of comparing the rating information and the lock menu for the corresponding program by the EIT during channel switching or transmission of a new program by the user, and if the new program is not locked (lock) in the program as it is A second step of stopping the output of video and audio when the lock is engaged; and a third step of determining whether the user is a registered user by accepting the user's voice after stopping the output of the video and audio, and the registered user. In this case, the stopped video and audio are output. If the user is not a registered user, the video and audio output is stopped.

Hereinafter, with reference to the accompanying drawings in detail as follows.

3 is a block diagram of a digital TV receiver having a speaker recognition module according to an embodiment of the present invention. As shown in this figure, a signal of a desired channel is input from an antenna, and a transport signal sequence obtained by detecting and error correcting the signal is shown. An output tuner 101, a packet separator 102 that receives the transport signal sequence, and separates the image signal sequence, an audio signal sequence, and an additional information signal sequence; (R), the video is an audio / video decoder 103 and 104 respectively output to the CRT, and receiving the additional information signal sequence and the database to the information that can be provided to the viewer to perform the speaker recognition A microcomputer 108 for controlling the computer, a user interface 112 for transmitting a user command, and a key input from the user interface 112. An OSD processor 107 for receiving OSD information from the ROM and outputting an OSD image, a switch 105 for switching the outputs of the video decoder 104 and the OSD processor 107, and the switch 105 An NTS encoder 106 for outputting the luminance signal (Y) and the color signal (C) so that the video signal transmitted through the general TV or monitor (CRT), and an audio input unit for inputting the user's voice And the user's voice feature extracted through the voice input unit 113 to find a model having the highest degree of similarity by comparing with a user who is registered in advance, and comparing it with a threshold value. It is composed of a speaker recognition module 114 to determine whether or not the recognition and deliver to the microcomputer (108).

FIG. 7 is a first embodiment of a speaker recognition method of a digital TV receiver having a speaker recognition module according to the present invention. As shown therein, a first step of muting audio when a user approaches a lock menu (S42) And a second step (S43) of accepting the user's voice after requesting the input of the user's voice after the audio mute, and a third step (S44) of canceling the audio mute after the user's voice is accepted. A fourth step (S45) of recognizing the speaker by using the characteristics of the user's voice received in the second step; a fifth step (S46) of searching whether the recognized speaker is a registered speaker; If the speaker is registered in the (6) (S46, S47) to allow the access to the lock menu, and not to access the lock menu unless the registered speaker.

FIG. 8 is a second embodiment of a speaker recognition method of a digital TV receiver having a speaker recognition module according to the present invention. As shown in FIG. 8, when a channel is switched by a user or a new program is transmitted, A first step (S51, S52) of comparing the rating information and the lock menu, and outputting the program as it is if the new program is not locked, and a second of muting video and audio if the lock is locked. Steps S53 and S54, a third step S56 of receiving a user's voice after performing video and audio muting and performing speaker recognition, and a video and audio mute if the recognized speaker is a registered speaker. In operation S58, the controller outputs the canceled output and stops output of video and audio unless the speaker is registered.

Referring to the operation and effect of the present invention configured as described in detail as follows.

When the tuner 101 receives a signal from an antenna and provides a detected and error-corrected transport signal sequence to the packet separator 102, the packet separator 102 demultiplexes the transport signal sequence to thereby transmit a video signal sequence and an audio signal. The signal string and the additional information signal string are separated and output.

Accordingly, the voice signal string is extended by the audio decoder 103 and then converted into an analog voice signal through an audio digital / analog converter 111 and provided to L and R of a speaker to output voice.

The video signal sequence is extended by the video decoder 104 and then output to the NTC encoder 106.

At this time, the microcomputer 108 reads the additional information separated by the packet separator 102, analyzes and combines the data into information that can be provided to the viewers, and stores the data in the RAM 110 and uses it when necessary. .

When the microcomputer 108 reads the OSD information corresponding to the key input from the user interface 112 into the ROM 109 and transmits the OSD information to the OSD processor 107, the OSD processor 107 transmits the OSD information. The image is processed as an OSD image and output to the switch 105.

In the digital TV receiver performing such an operation, the present invention intends to use speaker recognition instead of the password input method currently used as a general low.

When a user inputs a voice through the voice input unit 113, the speaker recognition module 114 receives the voice and performs the process illustrated in FIG. 4.

That is, first, the user's voice is input.

When the user's voice comes in, it is extracted as a feature that can well represent the user's voice characteristics while reducing the size of the voice.

The extracted voice features of the user and the models (parameters) of the registered users are compared.

In comparison, the similarity between the model of the user and the users registered in advance is calculated to find and select the most matching model (S14).

Then, by comparing the value of the model selected above with a predetermined threshold (S15), if it is higher than the threshold, it is recognized that the recognition is correct (S16), if it is lower than the threshold It is judged that the recognition is not correct, and makes an impossible decision (S17).

When the permission and non-determination is provided to the microcomputer 108, the microcomputer 108 performs the corresponding operation, which will be described later.

When shipping a digital TV receiver equipped with a speaker recognition module operating as described above to enable the user to register, this will be described based on FIG.

At the time of initial shipment, the lock is outputted so that the initial users who use the speaker recognition function can be registered (S21) and receive the voice of the initial user. (S22)

In response to the user's voice, according to the speaker recognition algorithm, the user's characteristics are well represented and a parameter (model) that can minimize the size is extracted and stored.

As described above, a process of extracting a parameter (model) is shown in FIG. 6. The method used for estimating the model uses a pattern matching method based on a statistical random process.

In other words, after initializing the model (S31), the REESTIMATION process is performed, which is a process of searching for a model that more closely matches the input value extracted from the feature. Examples of statistical re-estimation methods include the Baum-welch algorithm and the Segmental k-means algorithm. If the re-estimation process is terminated according to a specific criterion, the model is saved (S33).

In FIG. 5, the parameter representing the characteristic from the user's voice is estimated and stored (S23), and then locked (Lock on). (S24)

Thereafter, the user may be selected according to a program through the speaker recognition module 114.

Since the mode is applied only to the initial setting at the time of shipment from the factory mode, when all processing is completed, the mode is switched to the normal user mode.

Thereafter, speaker recognition is performed in two ways in the present invention with respect to the general user mode. First, a process of recognizing a speaker based on FIG. 7 and determining whether a lock menu is accessed according to whether the speaker is recognized will be described. Let's see.

First, when the user selects a key for accessing the lock menu through the user interface 112, the microcomputer 108 recognizing this is read into the ROM 109 storing the additional information and output to the OSD processor 107. To be displayed on the CRT through the NTC encoder 106.

Then, the microcomputer 108 mutes the audio to prevent contamination of the user's voice input by the TV speaker output (S42).

Afterwards, the user is asked to input voice through OSD or speaker.

When the user inputs the voice through the voice input unit 113 in this way, the microcomputer 108 transmits the voice to the speaker recognition module 114. (S43)

The audio is muted after receiving a voice input (S44).

The speaker recognition module 114 searches whether the speaker is registered through the process of FIG. 4 described above (S45), and transmits it to the microcomputer 108, and whether the microcomputer 108 is the registered speaker. (S46)

As a result of the determination, the registered speaker is granted permission to access the lock menu (S47), and if the registered speaker is not registered, the permission is not granted to access the lock menu.

However, only the operation of FIG. 7 makes the user feel uncomfortable.

For example, minor protection may require frequent changes to the lock menu, and the user may accidentally forget to change the menu, making it accessible to programs that minors should not watch.

For users who are burdened with frequent changes of the menu, the menu can turn off most grade values and use the operation shown in FIG. 8.

That is, when a program is newly started on an existing channel or the channel is switched by the user (S51), the rating information for the corresponding program by the EIT is compared with the value set in the lock menu (S52).

As a result of the comparison, if there is a lock, this is a program that should not be watched by minors. Therefore, the audio and video are muted to recognize the speaker, and the user's voice is requested (S54).

When the user inputs a voice through the voice input unit 113, the speaker recognition module 114 receives the voice. (S55)

Thus, the speaker recognition module 114 recognizes the speaker and searches whether the speaker is registered or not and informs the microcomputer 108 (S56).

Then, the microcomputer 108 outputs audio and video signals if the speaker is registered (S58), and outputs audio and video signals if the speaker is not a registered speaker.

Using the same method as described above can prevent minors from watching programs that they should not watch.

As described in detail above, the present invention provides a method for selecting a lock menu in a digital TV receiver by replacing speaker recognition with an existing password setting function, thereby making it safer, easier to manage, and easier to use.

Claims (4)

A packet separator which separates the transport signal sequence transmitted through the tuner into a video, audio, and additional information signal sequence, an audio / video decoder and an NTC encoder that extends the separated voice and image and outputs them to a speaker and a CRT; A TV receiver comprising an OSD processor for outputting OSD information corresponding to a key input from a user interface that transmits a user's command to an NC encoder, comprising: a voice input unit for inputting a user's voice and the voice input unit Speaker recognition module for searching whether or not the registered feature is extracted by comparing the feature value extracted from the user's voice and the registered speaker, and whether the audio and video signal is output according to the search result of the speaker recognition module. And a control unit for determining and performing the control operation accordingly. A digital television receiver comprising a speaker identification module, characterized by. A first step of accepting the user's voice after muting the audio when the user accesses the LOCK menu; a second step of accepting the user's voice and releasing the audio mute again; A third step of recognizing the speaker by extracting the voice feature; a fourth step of retrieving whether the recognized speaker is a pre-registered speaker; Or a fourth step of disallowing access to the lock menu, wherein the speaker recognition method of the digital TV receiver has a speaker recognition module. The method of claim 2, wherein the speaker recognition comprises: a first process of extracting a voice feature of a user; a second process of selecting a model having the highest degree of similarity by comparing the extracted feature value with a preregistered speaker; The third step of the digital TV receiver having a speaker recognition module, characterized in that the comparison process between the similarity of the selected model and the predetermined threshold value is granted only if the higher than the predetermined threshold value is properly recognized. Speaker recognition method. The first step of comparing the rating information and the lock menu of the corresponding program in the event information table (EIT) during channel switching or transmission of a new program by the user, and if the new program is not locked, A second step of muting the video and audio when the lock is locked, and a third step of performing speaker recognition by accepting the user's voice after the video and audio muting, and registering the recognized speaker. And a fourth step of outputting video and audio by canceling muting and stopping video and audio output if they are not registered.