KR20120083025A - Multimedia device for providing voice recognition service by using at least two of database and the method for controlling the same - Google Patents

Abstract

PURPOSE: A multimedia device for providing voice recognition service and a control method thereof are provided to enlarge the range of voice recognition. CONSTITUTION: A multimedia device extracts feature vector from received audio data(S620). In case the keyword corresponding to the extracted feature vector exists in the first database, the multimedia device controls the execution of a voice recognition service according to keyword(S630,S640). In case the corresponding key word does not exist, the multimedia device searches second database(S650). In case the corresponding key word exists in the second database, the multimedia device controls the execution of the voice recognition service according to the keyword(S660).

Description

MULTIMEDIA DEVICE FOR PROVIDING VOICE RECOGNITION SERVICE BY USING AT LEAST TWO OF DATABASE AND THE METHOD FOR CONTROLLING THE SAME}

The present invention relates to a multimedia device technology, and more particularly, to a multimedia device for providing a voice recognition service using at least two different databases and a control method thereof.

The image display device is, for example, a device having a function of receiving and processing a broadcast image that a user can watch. The image display device displays, for example, a broadcast selected by a user on a display among broadcast signals transmitted from a broadcasting station. Currently, broadcasting is shifting from analog broadcasting to digital broadcasting worldwide.

Digital broadcasting refers to broadcasting for transmitting digital video and audio signals. Digital broadcasting is more resistant to external noise than analog broadcasting, so it has less data loss, is advantageous for error correction, has a higher resolution, and provides a clearer picture. In addition, unlike analog broadcasting, digital broadcasting is capable of bidirectional services. On the other hand, smart TV (Smart TV) that combines the function of the video display device and the function of the multimedia device has recently been discussed.

Also, in recent years, some devices according to the prior art have introduced voice recognition technology. However, there is a limitation of using only a limited database, and thus there is a problem in that a wider range of speech recognition services cannot be provided to a user.

One embodiment of the present invention is to provide a solution that greatly increases the scope of speech recognition by using both an internal database and an external database of a multimedia device.

In addition, another embodiment of the present invention is to define a protocol for providing a service that a user wants to access more quickly by distinguishing a voice recognition command for controlling the multimedia device itself from a voice recognition command for providing keyword related data. .

In addition, another embodiment of the present invention is to provide a technique for ensuring voice recognition performance that is robust to external noise by using a mobile device that can communicate with a multimedia device.

According to an embodiment of the present invention, a method of controlling a multimedia device providing a voice recognition service using at least two different databases includes: receiving voice data of a user of the multimedia device, and receiving the received voice data. Extracting a feature vector required for recognition from the apparatus; determining whether a keyword corresponding to the extracted feature vector exists by using a first database located in the multimedia device; and presenting the feature vector in the first database. If the voice recognition service according to the keyword is executed, and if not present in the first database, using the second database located outside the multimedia device and connected to the network, the extracted feature vector The keyword corresponding to And determining whether the material and, if present in the second database, and a step of controlling so that the voice recognition services are running according to the keyword.

In addition, a multimedia device for providing a voice recognition service using at least two different databases (DataBase) according to an embodiment of the present invention, a voice sensor for receiving voice data of the user of the multimedia device A preprocessing unit for extracting a feature vector required for recognition from the received voice data, a recognition unit for analyzing the extracted feature vector using a first database located in the multimedia device, and the analysis result of the voice data The control unit controls to execute a device control command corresponding to the voice data when the tag includes a preset tag, and the extracted feature vector when the voice data does not include the preset tag as a result of the analysis. External, including the second database A network interface for sending in a vice.

According to one embodiment of the present invention, using both an internal database and an external database of a multimedia device, a solution for greatly increasing the scope of speech recognition is provided.

In addition, according to another embodiment of the present invention, by distinguishing between the speech recognition command for controlling the multimedia device itself and the speech recognition command for providing keyword-related data, a protocol for providing a service that the user wants to access more quickly is defined. .

In addition, according to another embodiment of the present invention, a technology for securing voice recognition performance that is robust to external noise using a mobile device capable of communicating with a multimedia device is provided.

More specific effects of the invention will be described later in detail in the table of contents.

1 is a view schematically showing an example of an entire system including a multimedia apparatus according to an embodiment of the present invention.
2 is a diagram illustrating a state in which a multimedia device, an external device, and a server are connected through a network according to an embodiment of the present invention.
3 is a diagram illustrating in detail a configuration module of a multimedia device according to an embodiment of the present invention.
4 is a diagram illustrating a first embodiment of an internal database (DataBase) for speech recognition of a multimedia device according to an embodiment of the present invention.
FIG. 5 is a diagram illustrating a second embodiment of an internal database (DataBase) for speech recognition of a multimedia device according to an embodiment of the present invention.
6 is a flowchart illustrating a method of controlling a multimedia device according to an embodiment of the present invention over time.
FIG. 7 is a flowchart illustrating in more detail a specific S640 step among all the steps shown in FIG. 6.
FIG. 8 is a flowchart illustrating in more detail a specific S660 step among all the steps shown in FIG. 6.
9 is a flowchart illustrating a control method of a multimedia device according to another embodiment of the present invention over time.
10 is a diagram illustrating an example of a voice recognition service of a multimedia device according to an embodiment of the present invention.
11 is a diagram illustrating another example of a voice recognition service of a multimedia device according to an embodiment of the present invention.
12 is a diagram illustrating another example of a voice recognition service of a multimedia device according to an embodiment of the present invention.

Hereinafter, various embodiments of the present invention will be described in detail with reference to the accompanying drawings. Furthermore, the suffixes "module" and "unit" for components used in the following description are merely given in consideration of ease of writing the present specification, and the "module" and "unit" may be used interchangeably with each other. Can be designed in hardware or software.

Meanwhile, the multimedia devices described herein correspond to various types of devices that receive and process broadcast data, for example. Furthermore, the multimedia device may correspond to a connected TV. The connected TV may include a wired / wireless communication device as well as a broadcast reception function, and thus may be more convenient to use, such as a handwritten input device, a touch screen, or a motion recognition remote controller. It can have In addition, by being connected to the Internet and a computer with the support of a wired or wireless Internet function, it is possible to perform functions such as e-mail, web browsing, banking or gaming. Standardized general-purpose operating systems may be used for these various functions.

Therefore, the connected TV can be freely added or deleted, for example, on a general-purpose OS kernel, so that various user-friendly functions can be performed. More specifically, the connected TV may be, for example, a web TV, an Internet TV, an HBBTV, a smart TV, a DTV, or the like, and may be applicable to a smartphone in some cases.

BRIEF DESCRIPTION OF THE DRAWINGS The above and other features and advantages of the present invention will be more apparent from the following detailed description taken in conjunction with the accompanying drawings, in which: FIG.

As used herein, terms used in the present invention are selected from general terms that are widely used in the present invention while taking into account the functions of the present invention, but these may vary depending on the intention or custom of a person skilled in the art or the emergence of new technologies. In addition, in certain cases, there is a term arbitrarily selected by the applicant, and in this case, the meaning will be described in the corresponding description of the invention. Therefore, it is intended that the terminology used herein should be interpreted based on the meaning of the term rather than on the name of the term, and on the entire contents of the specification.

1 is a diagram schematically showing an example of an entire broadcasting system including a multimedia apparatus according to an embodiment of the present invention. Although the multimedia apparatus of FIG. 1 may correspond to, for example, a connected TV, the scope of the present invention is not limited to the connected TV alone, and the scope of the present invention should be defined by the claims.

As shown in FIG. 1, an entire system including a multimedia apparatus according to an embodiment of the present invention includes a content provider (CP) 10, a service provider (SP) 20, and a network provider. (Network Provider; NP) (30) and HNED (40) can be divided. The HNED 40 corresponds to, for example, the client 100 which is a multimedia device according to an embodiment of the present invention.

The content provider 10 produces and provides various contents. As shown in FIG. 1, the content provider 10 includes a terrestrial broadcaster, a cable system operator or a multiple system operator, a satellite broadcaster, and an internet broadcaster. An internet broadcaster, etc. may be exemplified, etc. In addition, the content provider 10 may provide various applications and the like in addition to the broadcast content.

The service provider 20 may provide a service package of contents provided by the content provider 10. For example, the service provider 20 of FIG. 1 may package and provide a first terrestrial broadcast, a second terrestrial broadcast, a cable MSO, satellite broadcast, various internet broadcasts, applications, and the like to a user.

The network provider 30 may provide a network for providing a service to the client 100. The client 100 may establish a home network end user (HNED) to receive a service.

On the other hand, the client 100 can also provide content through the network. In this case, unlike the above, the client 100 may be a content provider, and the content provider 10 may receive content from the client 100. In the case of designing as described above, an interactive content service or a data service is possible.

2 is a diagram illustrating a state in which a multimedia device, an external device, and a server are connected through a network according to an embodiment of the present invention. Hereinafter, a process of using an internal database and an external database in a process of executing a voice recognition service by a multimedia device according to an embodiment of the present invention will be described below.

The multimedia device 200 according to an embodiment of the present invention is a device capable of speech recognition, and corresponds to, for example, a connected TV, a smart TV, a web TV, an Internet TV, a network TV, and the like. In addition, the multimedia device 200 includes an internal database 201 necessary for a voice recognition process. However, since the above-described internal database 201 includes only relatively limited data, the execution of functions according to voice recognition is also limited.

In order to solve such a problem, the multimedia device 200 according to an embodiment of the present invention is connected to the database 210 of the first device and the database 220 of the second device, which are located outside, The first device and the second device correspond to peripheral devices based on, for example, Digital Living Network Alliance (DLNA). Or, the first device and the second device described above correspond to, for example, USB, HDMI CEC connection equipment.

In addition, the multimedia device 200 is designed to communicate with a server 230 connected through a network such as the Internet so that an external database 231 may also be used in a voice recognition process.

Therefore, in this design, the multimedia device 200 does not rely only on the internal limited database 201, but can access external databases, thereby enabling a variety of voice recognition services. In addition, according to a specific condition, by designing to use an internal database or an external database selectively, there is an effect that the processing speed is also improved. Hereinafter, the operation of the multimedia device according to an embodiment of the present invention will be described in more detail with reference to FIG. 3.

3 is a diagram illustrating in detail a configuration module of a multimedia device according to an embodiment of the present invention. Hereinafter, referring to FIG. 3, a process of executing a voice recognition service by using a multimedia device according to an embodiment of the present invention using an internal or external database will be described in detail.

As shown in FIG. 3, the multimedia device 300 according to an embodiment of the present invention may include a voice sensor 301, a preprocessor 302, a recognizer 303, a controller 304, and a network interface 305. ), And the display unit 306 and the like. However, the modules shown in FIG. 3 are one embodiment, and the scope of the present invention should in principle be defined by the claims. In addition, although the voice sensor 301 inside the multimedia device 300 may be designed to detect a user's voice, the voice of the user may be detected using the mobile device 310 illustrated in FIG. 3. It is also possible. In this design, since the user can transmit voice data through the voice detecting sensor attached to the mobile device 310 which is closer to his or her mouth, the user may receive ambient noise or the audio signal of the multimedia device 300 itself. There is an advantage that can be excluded. In addition, the mobile device 310 that can communicate with the multimedia device 300 may be, for example, a mobile phone, a smartphone, a laptop, a tablet PC, or the like.

On the other hand, the speech recognition step of the multimedia device 300 according to an embodiment of the present invention can be largely divided into two areas. In other words, the preprocessor 302 and the recognizer 303 shown in FIG. 3 play a major role.

The preprocessor 302 extracts a feature vector required for recognition from the voice spoken by the user, and the recognizer 303 analyzes the feature vector to obtain a voice recognition result. For example, when a voice input through a microphone or the like enters the multimedia device 300, the preprocessor 302 may include feature vectors capable of expressing a phonetic feature every predetermined time (for example, 1/100 second). Extract.

The feature vectors described above well represent phonetic characteristics and should be insensitive to other factors such as background noise, speaker difference, and pronunciation attitude. Through this process, the recognition unit 303 focuses on purely phonetic characteristics and analyzes them. You can do it. The extracted speech feature vectors are passed to the recognition unit 303 and compared with the previously stored acoustic model, and the result is output as a final recognized sentence through language processing.

In particular, a database is used in the process of comparing with a previously stored acoustic model, the multimedia device according to an embodiment of the present invention is designed to be able to use the internal database and the external database at the same time. This will be described in detail later.

On the other hand, there are various types of feature vector extraction methods according to the point of view of mimicking how a human recognizes speech, and typically, LPC (Linear Predictive Coding) extraction method that analyzes with equal emphasis on all frequency bands or human Mel Frequency Cepstral Coefficients (MFCC) extraction, which reflects the characteristics of speech recognition that is not linear and follows Mel scale similar to log scale, high frequency emphasis extraction that emphasizes high frequency components to distinguish speech and noise clearly, The window function extraction method minimizes the distortion caused by breakage when analyzing the data into intervals.

The recognition unit 303 obtains a voice recognition result by comparing the phonetic information of a voice recognition related database located inside or outside the multimedia device 300 with a feature vector passed from the preprocessor 302. Done.

In addition, a database search process for speech recognition can be largely classified into a word unit search and a sentence unit search. The word unit search extracts the case of possible words by comparing the obtained feature vector with a word model stored in a database, that is, a phonetic feature of each word, or shorter. It is also called pattern classification because it is a process of finding an appropriate candidate pattern by comparing the pattern with the acoustic model stored in the database.

The result of the pattern classification process is passed to the sentence unit search in the form of a series of candidate words or candidate phonemes. In this process, based on the information of candidate words or candidate phonemes, it is judged which word or phoneme is most suitable by determining grammatical structure, overall sentence context, and conformity to a specific subject. For example, let's assume that 'i' goes' indistinguishable from '은' and '능' because of the indefinite pronunciation in the sentence.

At this time, the speech recognition system generates two candidate words as 'results' and 'performance' as a result of pattern classification. In the subsequent sentence unit search process, the sentence structure analysis reveals that '은' plays an investigation role in the sentence, and the investigation of 'performance' does not exist and thus is excluded from the candidate.

In other words, it is a process of improving the recognition performance through the restriction on the vocabulary and grammar structure. In this process, not only grammatical structure but also semantic information are used together, so it is also called language processing process. In addition, data used in pattern classification and language processing are learned by a computer in advance and stored in a database.

A representative technique mainly used in the recognition unit 303 is HMM (Hidden Markov Model), which is based on statistical pattern recognition, and combines a word unit search and a sentence unit search process into one optimization process. This method finds the best speech unit for this pattern by storing statistical information of patterns corresponding to the speech unit in the form of probability model and calculating the probability that the pattern can come from each model when an unknown input pattern comes in. to be.

Referring to FIG. 3 again, the recognition unit 303 accesses and recognizes a database required for speech recognition (particularly, according to an embodiment of the present invention, a hybrid DB such as an internal database and an external database) is used. The process of processing the result is described in detail as follows.

As shown in FIG. 3, the voice sensor 301 receives voice data of a user of the multimedia device 300, and the preprocessor 302 receives a feature vector required for recognition from the received voice data. Extract. Alternatively, the preprocessor 302 may be designed to receive voice data from the external mobile device 310.

The recognition unit 303 first analyzes the extracted feature vector using a first database located in the multimedia device 300. When the voice data includes a preset tag as a result of the analysis, the controller 304 controls the device control command corresponding to the voice data to be executed.

On the other hand, if the voice data does not include a predetermined tag as a result of the analysis, the network interface 305 transmits the extracted feature vector to an external device including a second database. The external device corresponds to, for example, the server 230 illustrated in FIG. 2.

Therefore, since the internal DB is used for the speech recognition database related to the keyword controlling the function of the multimedia device 300 itself, the command can be executed more quickly and the function itself is limited. Therefore, there is no need to use the external DB. Furthermore, since an external DB is used for the speech recognition database related to the keyword irrelevant to the function of the multimedia device 300, there are advantages in that various speech recognition services can be used. The external DB is distinguished from the internal DB in that it is easy to update and has a relatively large amount of data.

On the other hand, the predetermined tag described above is composed of a combination of specific letters for executing a command for controlling the multimedia device 300, for example. This will be described later in more detail with reference to FIGS. 4 and 5.

In addition, the multimedia device 300 according to another embodiment of the present invention further includes a display unit 306 as shown in FIG. 3. When the current state of the multimedia device 300 is a typing mode, the display unit 306 displays the keyword itself received from the external device described above in a typing area. In contrast, when the current state of the multimedia device 300 is not a typing mode, the display unit 306 is designed to display a content list related to a keyword received from the external device. The external device corresponds to, for example, a server (230 shown in FIG. 2) connected to the multimedia device 300 in a network. Meanwhile, the screen output by the display unit 306 will be described later in more detail with reference to FIGS. 10 to 12.

4 is a diagram illustrating a first embodiment of an internal database (DataBase) for speech recognition of a multimedia device according to an embodiment of the present invention. Hereinafter, referring to FIG. 4, a multimedia device according to an embodiment of the present invention will be described with reference to a method of performing a voice recognition service using an internal database storing a common tag of "TV".

As shown in FIG. 4, keywords that control the functions of multimedia devices (eg, TVs, etc.) themselves are commonly set to be tagged "TV" and stored in an internal database. Therefore, if the voice of "TV" is recognized, the internal database is quickly accessed. If a word other than "TV" is recognized, the internal or external DB can be accessed to collect more information. .

In addition, by recognizing the individual commands (volume up, volume down, power on, power off shown in FIG. 4) pronounced after "TV", and controlling to perform the corresponding control command, unnecessary access to the external database is performed. This saves time wasted in the process.

FIG. 5 is a diagram illustrating a second embodiment of an internal database (DataBase) for speech recognition of a multimedia device according to an embodiment of the present invention. Hereinafter, referring to FIG. 5, a multimedia device according to an embodiment of the present invention will be described with reference to a method of performing a voice recognition service using an internal database storing a common tag of “device”.

As illustrated in FIG. 5, keywords that control the functions of multimedia devices (eg, devices, etc.) themselves are commonly set to be tagged with "device" and stored in an internal database. Therefore, when the voice of "device" is recognized, it is designed to quickly access the internal database, and when a word other than "device" is recognized, it is designed to collect more information by accessing the internal DB or external DB. .

In addition, by recognizing the individual commands (channel up, channel down, power on, power off shown in Fig. 5), which are pronounced next to the "device," and then controlling to perform the corresponding control commands, unnecessary access to the external database is also possible. This saves time wasted in the process.

6 is a flowchart illustrating a method of controlling a multimedia device according to an embodiment of the present invention over time. Hereinafter, referring to FIG. 6, a method of providing a speech recognition service by selectively using an internal DB and an external DB according to an embodiment of the present invention will be described.

A multimedia device providing a voice recognition service using at least two different databases, receives voice data of a user of the multimedia device (S610). In addition, the multimedia device extracts a feature vector required for recognition from the received voice data (S620).

The multimedia device determines whether a keyword corresponding to the extracted feature vector exists using the first database located in the multimedia device (S630).

If the determination result (S630) exists in the first database, the multimedia device controls to execute a voice recognition service according to the keyword (S640).

On the other hand, if the determination result (S630) does not exist in the first database, the multimedia device is located outside the multimedia device and accesses a second database connected via a network (S650). In addition, the multimedia device determines whether a keyword corresponding to the extracted feature vector exists using the second database.

If present in the second database, the multimedia device controls to execute a voice recognition service according to the keyword (S660).

FIG. 7 is a flowchart illustrating in more detail a specific S640 step among all the steps shown in FIG. 6. Hereinafter, referring to FIG. 7, detailed steps of step S640 illustrated in FIG. 6 will be described.

The above-described step S640 further includes a step S641 of determining whether the keyword includes a predetermined tag as shown in FIG. 7. Since the tag has been described in detail with reference to FIGS. 4 and 5, the same description will be omitted.

In operation S640, when the determination result S641 is included, controlling the device control command corresponding to the keyword to be executed in operation S642, and when the determination result S641 is not included, The method may further include displaying the content list related to the keyword (S643). In particular, step S642 described above may be supplementarily interpreted with reference to FIG. 10, and step S643 described above may be supplementarily interpreted with reference to FIG. 11.

FIG. 8 is a flowchart illustrating in more detail a specific S660 step among all the steps shown in FIG. 6. Hereinafter, the detailed steps of the step S660 illustrated in FIG. 6 will be described with reference to FIG. 8.

The above-described step S660 may further include determining whether the current state of the multimedia device is a typing mode as shown in FIG. 8 (S661).

In operation S660, when the determination result S661 is a typing mode, displaying the keyword itself in a typing area in operation S662, and when the determination result is not a typing mode, is related to the keyword. The method may further include displaying the content list (S663). In particular, step S662 described above may be supplementarily interpreted with reference to FIG. 12, and step S663 described above may be supplementarily interpreted with reference to FIG. 11.

9 is a flowchart illustrating a control method of a multimedia device according to another embodiment of the present invention over time. Hereinafter, a method of controlling a multimedia device according to another exemplary embodiment of the present invention will be described by dividing the embedded area and the server area. In particular, the embedded area means, for example, a part for processing voice recognition using the internal database described above, and the server area means, for example, a part for processing voice recognition using the external database described above. Means.

The multimedia device according to another embodiment of the present invention receives voice data of a user (S910). Further, the multimedia device primarily determines whether the received voice data exists in an internal DB (S920). The internal DB is designed to include a common tag, for example, as shown in FIG. 4 or 5.

If the determination result (S920) exists in the internal DB, it is again determined whether the recognized voice data corresponds to the command (S930). If it corresponds to the determination result (S930) command, a specific function of the multimedia device is automatically executed (S940).

On the other hand, if it does not correspond to the determination result (S930) command, a search result related to the recognized voice data is displayed (S980).

On the other hand, if the determination result (S920) does not exist in the internal DB, the recognized voice data is transmitted to the external server (S950). At this time, it is designed to use the database of the external server.

In operation S960, it is determined whether the multimedia device is in a typing mode. In the determination result (S960) typing mode, a dictation is performed on the recognized speech data itself (S970). In the determination result (S960), the search result related to the recognized speech data is displayed. Display (S980).

In summary, when the speech recognition result shows that a common tag is inserted (for example, "TV on", "TV Off", "TV Volume up", etc.), the command is immediately responded to. Perform device control.

On the other hand, when the voice recognition result does not exist (for example, "TV"), the common separator ("Muan challenge", "boy age", etc.), the external server is accessed to perform a search.

In addition, when the word matching rate is low in the command DB (Embedded type) during speech recognition (for example, a word not present in the command list illustrated in FIG. 4 or 5), the voice data is transmitted through the server. Process voice data in the area.

On the other hand, when a search for the content of the voice recognition result is requested, the search is performed using a database of other devices (eg, PC, Set-top, Mobile, Network HDD) located in or near the multimedia device. In addition to the above-described process, it may be designed to simultaneously perform web search on a remote network by transmitting voice data to a server.

Then, the input state of the multimedia device is first checked to determine a user's intention as to whether a simple dictation is a purpose or a related content search.

10 is a diagram illustrating an example of a voice recognition service of a multimedia device according to an embodiment of the present invention. Hereinafter, a process of performing a specific function using an internal database among voice recognition services of a multimedia device according to an embodiment of the present invention will be described with reference to FIG. 10.

First, as shown in FIG. 10A, it is assumed that the multimedia device 1000 according to an embodiment of the present invention has a volume state set to a “3” level 1030. At this time, when the user 1010 of the multimedia device 1000 speaks 'TV volume up' 1020, the corresponding voice data is transmitted to the multimedia device 1000. At this time, the multimedia device 1000 is designed as shown in FIG.

Next, the multimedia device 1050 that recognizes the corresponding voice data using an internal database automatically raises the TV volume state by one step to set the "4" level 1060 as shown in FIG. do. The internal database has been described in detail with reference to FIGS. 4 and 5, and the same description will be omitted.

11 is a diagram illustrating another example of a voice recognition service of a multimedia device according to an embodiment of the present invention. Hereinafter, a process of providing a related content list using an external database among voice recognition services of a multimedia device according to an embodiment of the present invention will be described with reference to FIG. 11.

First, as illustrated in FIG. 11A, when the user 1110 of the multimedia device 1100 according to an embodiment of the present invention speaks as 'Mic Jacks' 1120, the corresponding voice data is Transmitted to the multimedia device 1100. In this case, the multimedia device 1100 is designed as shown in FIG. 3.

Next, the multimedia device 1150 that recognizes the corresponding voice data using an external database may include a content list 1160, 1170, 1180, and 1190 related to the recognized voice data as illustrated in FIG. 11B. ) Is designed to display automatically.

12 is a diagram illustrating another example of a voice recognition service of a multimedia device according to an embodiment of the present invention. Hereinafter, a process of executing dictation of voice data recognized using an external database among voice recognition services of a multimedia device according to an embodiment of the present invention will be described as follows.

First, as illustrated in (a) of FIG. 12, when the user 1210 of the multimedia device 1200 according to an embodiment of the present invention speaks as 'thickness method' 1220, corresponding voice data Is transmitted to the multimedia device 1200. At this time, the multimedia device 1200 is designed as shown in FIG.

Next, the multimedia device 1250 that recognizes the corresponding voice data by using an external database outputs the recognized voice data 1260 as shown in FIG. Unlike FIG. 11, the reason for not displaying the related content list is that the current multimedia device state corresponds to the keyword typing mode as shown in FIG. 12. Therefore, it would be a unique effect of the present invention to provide different kinds of results according to the current mode and state of the multimedia device.

According to the embodiments of the present invention described above, when the command execution using the internal database maintains a fast execution time, and when a large amount of data processing is required, a more dynamic content search result is displayed using the external database. There is an advantage.

On the other hand, in the present specification, both the object invention and the method invention are described, and the description of the invention can be supplementally applied as necessary. In addition, for convenience of description, each drawing is divided and described, but implementing another embodiment by combining the drawings or embodiments is also within the scope of the present invention.

Method invention according to the present invention are all implemented in the form of program instructions that can be executed by various computer means can be recorded on a computer readable medium. The computer readable medium may include program instructions, data files, data structures, etc. alone or in combination. The program instructions recorded on the medium may be those specially designed and constructed for the present invention or may be available to those skilled in the art of computer software. Examples of computer-readable recording media include magnetic media such as hard disks, floppy disks, and magnetic tape, optical media such as CD-ROMs, DVDs, and magnetic disks, such as floppy disks. Magneto-optical media, and hardware devices specifically configured to store and execute program instructions, such as ROM, RAM, flash memory, and the like. Examples of program instructions include not only machine code generated by a compiler, but also high-level language code that can be executed by a computer using an interpreter or the like. The hardware device described above may be configured to operate as one or more software modules to perform the operations of the present invention, and vice versa.

As described above, the present invention has been described by way of limited embodiments and drawings, but the present invention is not limited to the above embodiments, and those skilled in the art to which the present invention pertains various modifications and variations from such descriptions. This is possible.

Therefore, the scope of the present invention should not be limited to the described embodiments, but should be determined not only by the claims below but also by the equivalents of the claims.

200: multimedia device
210: first device database (DB)
220: second device database (DB)
230: server

Claims (11)

In the control method of a multimedia device providing a voice recognition service using at least two different databases,
Receiving voice data of a user of the multimedia device;
Extracting a feature vector required for recognition from the received speech data;
Determining whether a keyword corresponding to the extracted feature vector exists by using a first database located in the multimedia device;
If present in the first database, controlling a voice recognition service according to the keyword to be executed;
Determining whether a keyword corresponding to the extracted feature vector exists by using a second database located outside the multimedia device and connected to a network when the data does not exist in the first database; And
Controlling the voice recognition service according to the keyword to be executed when the second database exists in the second database.
Control method of a multimedia device comprising a. The method of claim 1,
When present in the first database, the step of controlling the voice recognition service according to the keyword is executed,
Determining whether the keyword includes a predetermined tag;
If the determination result is included, controlling a device control command corresponding to the keyword to be executed; And
If it is not included in the determination result, displaying a content list related to the keyword.
Control method of a multimedia device comprising a. The method of claim 1,
When present in the second database, the step of controlling to execute the voice recognition service according to the keyword,
Determining whether a current state of the multimedia device is in a typing mode;
If the determination result is a typing mode, displaying the keyword itself in a typing area; And
If the determination result is not a typing mode, displaying a content list related to the keyword;
Control method of a multimedia device comprising a. The method of claim 1,
The second database,
And a database managed by a server connected to the multimedia device via a network. The method of claim 1,
Wherein the receiving comprises:
And receiving voice data obtained from a voice detecting sensor attached to a mobile device capable of communicating with the multimedia device. A computer-readable recording medium in which a program for executing the method of any one of claims 1 to 5 is recorded. In the multimedia device that provides a speech recognition service using at least two different databases,
A voice sensor for receiving voice data of a user of the multimedia device;
A preprocessor extracting a feature vector required for recognition from the received speech data;
A recognizer configured to analyze the extracted feature vector using a first database located in the multimedia device;
A controller for controlling a device control command corresponding to the voice data to be executed when the voice data includes a predetermined tag as a result of the analysis; And
If the voice data does not include a predetermined tag as a result of the analysis, a network interface for transmitting the extracted feature vector to an external device including a second database
Multimedia device comprising a. The method of claim 7, wherein
If the current state of the multimedia device is a typing mode, display the keyword itself received from the external device in a typing area, and
A display unit configured to display a content list related to a keyword received from the external device when the current state of the multimedia device is not a typing mode;
Multimedia device further comprising. The method of claim 8,
The external device,
And a server connected to a network connected to the multimedia device. The method of claim 7, wherein
The preset tag,
And a specific letter combination for executing the device control command. The method of claim 7, wherein
The multimedia device,
A multimedia device corresponding to at least one of a network TV, a connected TV, a smart TV, a web TV, and an Internet TV.

Images