JP2019215393A - Image display device and television receiver - Google Patents

Abstract

To provide technology for changing a voice recognition range to be suitable to a viewing stye of a user in a display device with a voice recognition function.SOLUTION: A display device with a voice recognition function, having a display section, comprises a voice input section which: has directivity, sets an orientation direction being a direction of directivity to a prescribed angle, performs voice acquisition in parallel for a plurality of times, in which angles of voice recognition differ, which selects candidate voice data in accordance with a feature amount of voice data from voice data acquired by the voice acquisition, sets the orientation direction of the candidate voice data to be a speaker direction, discriminates whether or not valid voice data is a keyword voice, performs the voice acquisition in which the speaker direction is set to be the orientation direction and acquires a command voice. The display device comprises: a narrow angle mode in which a voice acquisition range by the voice acquisition for the plurality of times is narrow; and a wide angle mode in which the voice acquisition range by the voice acquisition for the plurality of times is wide, and the display device is configured to allow the narrow angle mode and the wide angle mode to be switched.SELECTED DRAWING: Figure 1

Description

  The present invention relates to a television with a voice recognition function. More specifically, the present invention relates to improving the accuracy of speech recognition in a television with a speech recognition function.

  In recent years, many products equipped with a voice recognition function have been released. The voice recognition function is convenient because the device can be operated and necessary information can be easily obtained without using a remote controller or the like. Products equipped with a voice recognition function on a television are also on the market.

  When using the voice recognition function on the TV, there are cases where the volume of the TV is raised or lowered, a channel is switched, a remote controller is replaced, or information such as tomorrow's weather is checked. In any case, it is generally considered that the user faces the television and is in a position suitable for watching the television. Therefore, it is considered appropriate that the direction of voice recognition be only the direction directly facing the television. By suppressing noise in other directions, the speech recognition rate can be increased.

However, this is not always the case depending on the viewing style of the user. For example, a user may be watching television in the living room from the kitchen. In some cases, the user watches television as background music while performing other tasks.
Therefore, there has been a demand for a technique capable of changing a voice recognition range in accordance with a user's viewing style.

  To solve such a problem, various techniques have been conventionally proposed. For example, a technique has been disclosed in which the position of a viewer is captured by a camera, and the direction of the directivity of voice recognition is changed (Japanese Patent Application Laid-Open No. H10-163,086).

Patent No. 6250297

  However, setting of directivity is difficult for a plurality of viewers, and setting is difficult when the camera image is unclear. Therefore, the problem of changing the voice recognition range in accordance with the user's viewing style has not been solved.

  The image display device according to the present invention has directivity, sets a directivity direction, which is a direction of directivity, to a predetermined angle, performs sound acquisition a plurality of times in parallel, and the angle of each sound acquisition is different. Selecting candidate voice data from the voice data acquired by the voice acquisition in accordance with the feature amount of the voice data, setting the direction of the candidate voice data to the speaker direction, and determining whether the valid voice data is the keyword voice. A voice input unit that obtains the command voice by performing the voice acquisition with the speaker direction as the directivity direction, and a narrow angle mode in which the voice acquisition range by the voice acquisition is narrow; A wide-angle mode in which the voice acquisition range by the voice acquisition is wide, and the mode can be switched between the narrow-angle mode and the wide-angle mode.

According to the present invention, since the voice recognition range can be changed according to the user's viewing style,
The voice recognition rate can be improved.

It is a block diagram of a display concerning the present invention. FIG. 2 is a diagram illustrating one embodiment of a display device according to the present invention. 5 is a flowchart illustrating an embodiment of a display device according to the present invention. It is a figure showing the menu display of the display concerning the present invention. FIG. 4 is a diagram illustrating a screen display of the display device according to the present invention. FIG. 11 is a diagram illustrating another embodiment of the display device according to the present invention. 9 is a flowchart illustrating another embodiment of the display device according to the present invention.

The greatest feature of the image display device of the present invention is that the voice recognition range can be changed in accordance with the viewing style of the user. Hereinafter, embodiments will be described with reference to the drawings.
Note that the overall shape and the shape of each part of the image display device shown in the present embodiment are not limited to the embodiments described below, and are within the scope of the technical idea of the present invention, that is, the same operation and effect. It can be changed within the range of shapes and dimensions that can be exhibited.
(Example 1)

The configuration of the present invention will be described with reference to FIGS.
FIG. 1 is a block diagram of a display device according to the present invention. FIG. 2A is a diagram illustrating a narrow angle mode in one embodiment of the display device according to the present invention. FIG. 2B is a diagram illustrating the wide angle mode. FIG. 3 is a flowchart illustrating an embodiment of the display device according to the present invention. FIG. 4A is a diagram showing a menu display of the display device according to the present invention.

(Explanation of block diagram)
The display device 1 is a television receiver with a voice recognition function, and is a general television receiver equipped with a voice recognition function.
The display device 1 mainly includes an audio input function, a control function, a video display function, and an audio output function.

The voice input function includes a microphone array 100 and a voice input processing unit 200. The microphone array 100 includes a plurality of omnidirectional microphones. In this example, there are four microphones. The microphones are arranged in a straight line. The sound that enters each microphone is delayed and added for each microphone in the subsequent stage, and is used as sound information having directivity. The directional direction can be changed by changing the delay amount. This method is also called beamforming. This operation is performed in parallel, and a plurality of pieces of audio information are generated.
In other words, a plurality of voice acquisitions having directivity and capable of setting the directivity direction are performed in parallel.
By making the directions of obtaining a plurality of sounds different, it is possible to obtain a wide range of sounds that cannot be covered by one sound obtaining operation.
In addition, for one audio acquisition, by giving directivity to the normal direction to the display surface of the display unit of the television, that is, the front direction of the television, it is possible to acquire audio with emphasis on the front of the television.
The voice input processing unit 200 is a part that performs processing on voice input by the microphone array 100, and performs noise removal of voice, recognition of keywords, determination of command voice, and the like. The voice input processing unit 200 mainly includes a preprocessing unit 210, a voice recognition unit 220, and a voice recognition parameter unit 230.

  The preprocessing unit 210 is a unit that removes noise, echo components, and the like in order to increase the recognition rate of voice input from a microphone. For example, echo cancellation and noise reduction are performed. In the echo cancellation, a sound component emitted from the television is removed using the sound information from the sound processing unit 600. By this processing, recognition failure and erroneous recognition due to television sound can be removed. In the noise reduction, low-frequency and high-frequency noises are removed, and stationary noise is removed.

The voice recognition unit 220 performs two operations of keyword detection and command detection.
At the time of keyword detection, voice data for each of a plurality of angles is generated based on the plurality of voice data generated by the preprocessing unit 210. The valid voice data is specified according to the feature amount of the voice data. The directional direction at the time of acquiring the effective voice data is defined as the speaker direction. Then, it is determined whether the valid voice data is a keyword voice. As the feature amount of the voice data, the magnitude of the utterance volume can be considered. This is because the volume increases when the direction of the speaker matches the pointing direction.
If the utterance is a keyword voice, the control unit 300 is notified that the keyword has been received.
When a command is detected, a voice at a set angle is received, and voice data is transmitted to the control unit 300 as command voice data.
The voice recognition parameter section 230 is a parameter group for setting a voice receiving angle at the time of keyword detection. By changing this value, the range of voice acquisition or the direction of voice acquisition can be changed. In response to an instruction from the control unit 300, an appropriate parameter from the speech recognition parameter unit 230 is sent to the speech recognition unit 220.

The control function is a part that controls the entire display device 1. It mainly includes a control unit 300, a communication unit 310, and an RC light receiving unit 320.
The control unit 300 is a part that performs control of video and audio of the television, control related to voice recognition, and communication with the outside. It includes a microcomputer, and ROMs and RAMs as memories.
A voice recognition angle parameter is sent to the voice input processing unit 200 via the voice recognition parameter unit 230 to change the angle.
The command voice acquired by the voice input processing unit 200 is transmitted to the voice analysis unit 800 via the communication unit 310 and the Internet 700, and the analysis result from the voice analysis unit 800 is received as text information via the communication unit 310. Or The text of the result of the voice analysis unit 800 is displayed via the image processing unit 430, and the text of the result of the voice analysis unit 800 is converted into voice and transmitted to the voice processing unit 600 as voice data.

The communication unit 310 is a part that performs network communication with the outside. The user can connect to the Internet and exchange data with the voice analysis unit 800 on the Internet. It may be wired or wireless.
RC light receiving section 320 is a receiving section of a remote control of a television, receives data from an infrared remote control, and transmits the data to control section 300.

The video display function is related to the video of a television, and performs reception of broadcast waves and superimposition of text data from the control unit 300 on video. It mainly comprises a tuner 400, an antenna 410, a decoding unit 420, an image processing unit 430, and a display unit 500.
The tuner 400 is a tuner for general broadcast reception, and receives a broadcast wave from the antenna 410 and selects a program. The decoding unit 420 is a unit that converts the received data into TS and separates video, audio, meta information, and the like. The video is sent to the image processing unit 430 and the audio is sent to the audio processing unit 600.
The control unit 300 transmits the timing of the video / audio source to the video processing unit 430 and the audio processing unit 600. In addition, it creates a sub-screen that displays information such as effects of user operations and subtitles at appropriate positions, and outputs the sub-screen to the video processing unit. The sub-screen is a layer for superimposing on a video. In addition, a mark indicating that the voice recognition mode has been set when a voice input for information is performed, a text of a voice recognition result, a recognition range mode of voice recognition, and the like are displayed.
The image processing unit 430 performs adjustment of luminance, saturation, hue, and the like of an image, high definition and scaling of the image, and the like. In addition, a process of superimposing a sub screen or the like on a video is performed.

The audio output function is a part related to the audio output of the television. It mainly comprises an audio processing unit 600 and a speaker 610. The audio processing unit 600 generates an audio waveform from audio data using a DA converter or the like, and outputs audio to a speaker 610 via an amplifier.
In addition, the audio data is sent to the pre-processing unit 210 to cancel the echo. Data corresponding to the volume emitted from the speaker 610 or a volume level is added to the audio data to be sent to the preprocessing unit 210.

(Operation example in narrow angle mode)
Next, the operation in the narrow angle mode and the wide angle mode will be described with reference to FIGS.
FIGS. 2A and 2B show an operation example in the narrow angle mode. 2C and 2D are operation examples in the wide angle mode. FIG. 3 is a flowchart including both cases.
As an operation example, there are five voice detection directions, and each direction has a fan-shaped range with a voice detection angle of 60 degrees. The voice obtainable range is B1, B2, B3, B4, and B5 as shown in FIG. The direction of B3 corresponding to the center is defined as a voice acquisition area center direction C. In this example, the direction of the audio acquisition area center direction C is a direction facing the display surface of the television.
The step angle of each audio obtainable range is 10 degrees in the narrow angle mode and 30 degrees in the wide angle mode.

First, the operation in the narrow angle mode will be described with reference to FIGS. 2 (a) and 2 (b). The display device 1 has a shape viewed from the top, and the left side is the image display side and the right side is the back side of the television on the paper. The microphone array 100 is arranged at the center of the television image display position. Unless otherwise specified, the right and left sides are defined as the right side (direction of B5 on the paper) and the left side (direction of B1 on the paper) with respect to the user viewing the television from the front.
After starting the television, the voice recognition parameters are set (S101). The number of search directions is set to 5, the search direction is set to the front, and the search step angle is set to 10 degrees.
The keyword is “OK bagel”.

The user H speaks "OK bagel" (FIG. 2A). The microphone array 100 and the voice input processing unit 200 acquire voice in parallel for each of the five directions (S102).
The user H is located at a position slightly shifted leftward from the center direction C of the voice acquisition area. Since the direction of the loudest volume is the direction of the voice obtainable range B2, the speaker direction is set to B2, and voice data obtained from this direction is set as candidate voice data.
B2 is stored as the direction of the user H (S103).

  It confirms whether the utterance content is a keyword voice. If the keyword is confirmed, the command voice acquisition state is entered. If the keyword is not the keyword voice, the process returns to the voice acquisition operation (S102).

  First, as the command voice acquisition state, the voice acquisition direction is set to B2, which is the direction stored in S103 (FIG. 2B). Since the step angle in the voice acquisition direction is small, fine adjustment is possible. The position of the user H can be adjusted to almost the center of the voice acquisition range.

The process enters a command voice acquisition wait state (S105). When the user H speaks "Turn down the volume", the voice is acquired (S106, S107). The acquired audio data is sent to the audio analysis unit 800 via the control unit 300 and the communication unit 310 (S108). The voice analysis unit 800 analyzes the voice data using a database, a large computer, or the like, and converts the voice to text. In this example, it is “Onryousagete”. Since the sound is the content for instructing the operation of the television, “VOL DWN” which is a command of the television operation is generated. If the voice is the content of the inquiry, an answer to the inquiry is generated as text data.
The voice analysis unit 800 transmits to the display device 1 “Onkyo Sagete” which is data in which the command voice is converted into text and “VOL DWN” which is data as an answer (S109).

  The control unit 300 sends the video of the sub-screen displaying the text of the command voice to the image processing unit 430, and the image processing unit 430 superimposes the video of the sub-screen on the main video. If the answer data is TV operation data, a TV operation is performed. If the answer data is text data, a video of the sub screen of the answer is created, sent to the image processing unit 430, and the video of the sub screen is superimposed on the main video. At the same time, the control unit 300 generates audio data from the text, and transmits the audio data to the audio processing unit 600. The voice processing unit 600 outputs the text content by voice using the speaker 610 (S110).

  Through a series of operations, the direction of the user H near the front of the television can be specified, the voice acquisition angle can be adjusted, and the recognition rate of the utterance of the user H can be increased. Therefore, the narrow angle mode is optimal for a lifestyle user whose user is near the front of the television.

  In the description of the narrow angle mode of the present embodiment, when acquiring the keyword voice, a plurality of voices are obtained.However, only the voice is obtained with the direction normal to the display surface of the display unit of the television as the directivity direction. The command voice may be obtained in the same front direction. When the user is in front of the television, the keyword voice and the command voice can be obtained without any problem.

  Also, when acquiring the keyword voice, similarly to the wide angle mode, a wide range of voice may be obtained, and when obtaining the command voice, only the frontal voice may be obtained. By doing so, it becomes easier to acquire the keyword voice.

(Operation example in wide angle mode)
Next, the operation in the wide angle mode will be described with reference to FIGS. 2C and 2D.
After starting the television, the voice recognition parameters are set (S101). The number of search directions is set to 5, the search direction is set to the front, and the search step angle is set to 30 degrees. The five sound acquisition directions cover almost 180 degrees in the direction of the display surface of the television. The position of the user H is set to be larger than the front of the television and to the left.

The user H speaks "OK bagel" (FIG. 2C). The microphone array 100 and the voice input processing unit 200 acquire voice in parallel for each of the five directions (S102).
The user H is located at a position shifted to the left more than the center direction C of the voice acquisition area. Since the direction of the loudest volume is the direction of the voice obtainable range B2, the speaker direction is set to B2, and voice data obtained from this direction is set as candidate voice data. B2 is stored as the direction of the user H (S103).

  It confirms whether the utterance content is a keyword voice. If the keyword is confirmed, the command voice acquisition state is entered. If the keyword is not the keyword voice, the process returns to the voice acquisition operation (S102).

  First, as the command voice acquisition state, the voice acquisition direction is set to B2, which is the direction stored in S103 (FIG. 2D). Since the step angle in the voice acquisition direction is large, it is possible to cover even if the position of the user H is largely shifted. However, the position of the user H may be shifted from the center of the search range. In that case, the voice recognition rate is low.

The process enters a command voice acquisition wait state (S105). When the user H speaks "Turn down the volume", the voice is acquired (S106, S107). The acquired audio data is sent to the audio analysis unit 800 via the control unit 300 and the communication unit 310 (S108). The voice analysis unit 800 analyzes the voice data using a database, a large computer, or the like, and converts the voice to text. In this example, it is “Onryousagete”. Since the sound is the content for instructing the operation of the television, “VOL DWN” which is a command of the television operation is generated. If the voice is the content of the inquiry, an answer to the inquiry is generated as text data.
The voice analysis unit 800 transmits to the display device 1 “Onkyo Sagete” which is data in which the command voice is converted into text and “VOL DWN” which is data as an answer (S109).

  The control unit 300 sends the video of the sub screen displaying the text of the command voice to the image processing unit 430, and the image processing unit 430 superimposes the video of the sub screen on the main video. The sub screen is a graphic plane that draws a content in which other image data or the like is superimposed on the main video. If the answer data is TV operation data, a TV operation is performed. If the answer data is text data, a video of the sub screen of the answer is created, sent to the image processing unit 430, and the video of the sub screen is superimposed on the main video. At the same time, the control unit 300 generates audio data from the text, and transmits the audio data to the audio processing unit 600. The voice processing unit 600 outputs the text content by voice using the speaker 610 (S110).

  Through a series of operations, it is possible to identify the direction of the user H located at a position significantly deviated from the front of the television, adjust the voice acquisition angle, and recognize the utterance of the user H. Therefore, the wide angle mode is most suitable for a lifestyle user who is often not near the front of the television.

(Menu display)
As described above, the voice recognition mode of the display device 1 needs to have at least the narrow angle mode and the wide angle mode. Therefore, a menu is used in which the user can easily switch between the two modes (FIG. 4A). A menu screen 510 is displayed on the display section 500 of the display device 1. It has a narrow angle mode selection button 520 and a wide angle mode selection button 530. The two buttons are selected and determined by, for example, an up / down key and a determination button on a remote controller. FIG. 2A shows a screen in which the wide angle mode is selected. When the wide angle mode is selected, the control unit 300 selects a wide angle mode parameter from the parameters of the voice recognition parameter unit 230 and sets the parameters of the voice recognition unit 220.

(Indication of voice recognition angle, etc.)
The display of the voice recognition angle will be described with reference to FIG. If there are multiple voice recognition modes, the user may need to confirm which mode is currently selected. FIG. 5A shows a case where the wide angle mode is set. A content video 550 is displayed on the display unit 500, and a voice recognition angle display area 560 is assigned to a lower part of the screen. In order to represent the wide angle mode, the effective voice recognition angle 570 is displayed as an elongated bar. The user sees the elongated bar-shaped display and recognizes the mode to recognize a wide angle.

  FIG. 5B shows a case where the narrow angle mode is set. A content video 550 is displayed on the display unit 500, and a voice recognition angle display area 560 is assigned to a lower part of the screen. To represent the narrow angle mode, the effective voice recognition angle 570 is displayed as a small bar. The user sees the small bar-shaped display and recognizes the mode of recognizing a narrow angle.

FIG. 5C is a diagram showing additional information. The voice acquisition display 580 is a display indicating that keyword voice and command voice are being acquired. Voice recognition angle display area 5
By displaying the display in the upper center portion of 60, the user can confirm that sound has been acquired. In addition, by displaying voice text 590, which is a text of voice input and answer, on the screen, the user can surely grasp the uttered content and the content of the answer.

  As described above, according to the present invention, in a television having a voice recognition function, the voice recognition range having a plurality of voice recognition ranges can be used to change the voice recognition range in accordance with the viewing style of the user. Convenience can be improved.

In other words, in other words, the configuration according to the present embodiment is a display device with a voice recognition function having a display unit, which has directivity, sets a directivity direction that is a directivity direction to a predetermined angle, and performs voice acquisition in parallel. And do it multiple times. The predetermined angle refers to, for example, a fixed step angle for each voice acquisition. The candidate voice data is selected according to the feature amount of the voice data, the direction of the candidate voice data is set as the speaker direction, and it is determined whether or not the candidate voice data is a keyword voice.
A voice input unit is provided for obtaining the voice with the utterer direction, which is the angle at which the candidate voice data was obtained, as the directional direction, and obtaining a command voice.
A narrow-angle mode in which a plurality of audio acquisition ranges are narrow (a narrow acquisition angle in the horizontal direction); and a wide-angle mode in which the plurality of audio acquisition ranges are wide (a narrow acquisition angle in the horizontal direction). , And can be switched between the narrow angle mode and the wide angle mode.

Also, in the display device with a voice recognition function, when the normal direction to the display surface of the display unit is set to the front direction, the directivity direction, which is the direction of the directivity, is set to the front direction, and voice acquisition is performed. Is determined as to whether or not the voice data acquired is a keyword voice. Voice acquisition is performed only in one direction.
The pointing direction is set to the front direction, the voice is obtained, and the command voice is obtained.
This mode is referred to as a narrow angle mode.
It has directivity, and sets the directivity direction, which is the direction of the directivity, to a predetermined angle, performs voice acquisition a plurality of times in parallel, the angle of each voice acquisition is different, and the voice data acquired by the voice acquisition is different. , Candidate voice data is selected in accordance with the feature amount of the voice data, the direction of the candidate voice data is set as the speaker direction, and it is determined whether or not the valid voice data is a keyword voice.
The voice is acquired with the speaker direction as the directional direction, and a command voice is acquired. This mode is referred to as a wide angle mode.
A narrow angle mode in which a plurality of voice acquisition ranges are narrow (the horizontal acquisition angle is narrow); and a wide angle mode in which the plurality of voice acquisition ranges are wide (the horizontal acquisition angle is narrow). , And can be switched between the narrow angle mode and the wide angle mode.

(Example 2)
In the first embodiment, an example has been described in which a plurality of voice recognition ranges are switched using a menu. However, displaying and setting a menu every time the two modes are changed may be troublesome in some cases.
Therefore, a technique for changing the mode without changing the setting in the menu has been demanded.

This will be described with reference to FIGS. 4B, 6, and 7. FIG. FIG. 4B is a menu display example of the present embodiment. 6A and 6B show the first speech recognition operation when the power is turned on. FIGS. 6C and 6D show the second and subsequent speech recognition operations. FIG. 7 is a flowchart of the first and second power-on speech recognition.
The mode of this embodiment will be described as an auto mode.

The menu shown in FIG. 4B is displayed, and the auto mode selection button 540 is selected and executed to set the mode to the auto mode.
After turning on the television, the voice recognition parameters are set to the wide angle mode (S201). As shown in FIG. 6A, a mode is set in which a wide range of sound is picked up.
The user H speaks "OK bagel" and acquires voice data (S202). The voice data in the five directions are compared, the direction with the highest volume is specified, and stored in the variable SCN1. In the example of FIG. 6A, B2 is stored in the variable SCN1 (S203).
If the voice is not a keyword, the process returns to S202. If it is a keyword, the process proceeds to S205 (S204). The voice acquisition direction is set to B2 stored in the variable SCN1 (S205). It waits for the command voice and acquires the voice (S206).
The data is transmitted to the voice analysis unit 800, the analysis result is received, and the control unit 300 displays the result on the display unit 500 or the like (S207).

The parameters are set in preparation for the second speech recognition. The mode is set to the narrow angle mode, and the center direction is the direction of B2 stored in the SCN1 (S208).
The direction of the sound acquisition area center direction C is not the front of the television but substantially the direction of the user H.
The user H speaks "OK bagel" and acquires voice data (S209). The sound data in the five directions are compared, the direction in which the volume is the highest is specified, and the direction is stored in the variable SCN2. In the example of FIG. 6C, B2 is stored in the variable SCN2 (S210).
If the voice is not a keyword, the process returns to S212. If it is a keyword, the process proceeds to S215 (S211). The voice acquisition direction is set to B2 stored in the variable SCN2 (S212). It waits for the command voice and acquires the voice (S213).
The data is transmitted to the voice analysis unit 800, the analysis result is received, and the control unit 300 displays the result on the display unit 500 or the like (S214).

With this operation, the voice recognition range can be accurately adjusted to the direction of the user even when the user is not in the front direction. The user can improve the speech recognition rate without making settings using the menu.
Further, even when the user is far away from the front of the television, the direction of the user can be accurately grasped, so that the recognition rate of voice recognition can be improved.

1 display device 100 microphone array 200 voice input processing unit 210 preprocessing unit 220 voice recognition unit 230 voice recognition parameter unit 300 control unit 310 communication unit 320 RC light receiving unit 400 tuner 410 antenna 420 decoding unit 430 image processing unit 500 display unit 510 Menu screen 520 Narrow angle mode selection button 530 Wide angle mode selection button 540 Auto mode selection button 550 Content video 560 Voice recognition angle display area 570 Voice recognition effective angle 580 Voice acquisition display 590 Voice text 600 Voice processing unit 610 Speaker 700 Internet 800 Voice Analysis unit H Users B1,..., B5 Voice acquisition possible range C Voice acquisition area center direction

Claims (8)

A display device with a voice recognition function having a display unit,
With directivity, set the directivity direction, which is the direction of the directivity, to a predetermined angle, perform voice acquisition multiple times in parallel,
The angle of each of the audio acquisitions is different,
From the voice data acquired by the voice acquisition, candidate voice data is selected according to the feature amount of the voice data, the direction of the candidate voice data is set to the speaker direction, and whether or not the candidate voice data is a keyword voice And determine
The voice acquisition unit performs the voice acquisition with the speaker direction as the directivity direction, and includes a voice input unit that acquires a command voice.
A narrow angle mode in which the voice acquisition range by the plurality of voice acquisitions is narrow;
A wide angle mode in which the voice acquisition range is wide by a plurality of the voice acquisitions,
A display device capable of switching between the narrow angle mode and the wide angle mode. A display device with a voice recognition function,
When the normal direction to the display surface of the display unit is set to the front direction, the directivity direction, which is the direction of the directivity, is set to the front direction and voice acquisition is performed, and whether or not the voice data acquired by the voice acquisition is a keyword voice is determined. Determine,
A narrow-angle mode in which the pointing direction is set to the front direction and the voice is obtained, and a command voice is obtained;
With directivity, set the directivity direction, which is the direction of the directivity, to a predetermined angle, perform voice acquisition multiple times in parallel,
The angle of each of the audio acquisitions is different,
From the voice data acquired by the voice acquisition, candidate voice data is selected according to the feature amount of the voice data, the direction of the candidate voice data is set to the speaker direction, and whether or not the candidate voice data is a keyword voice And determine
A wide-angle mode for performing the voice acquisition with the speaker direction as the directivity direction and acquiring a command voice,
A display device capable of switching between the narrow angle mode in which the sound acquisition range is narrow and the wide angle mode in which the sound acquisition range is wide.   The display device according to claim 1, wherein the feature amount of the audio data is a magnitude of an utterance volume.   4. The display device according to claim 1, wherein the switching of the mode is performed by selecting an item of a menu.   The display device according to claim 1, wherein a display according to the mode being set is performed on the display unit.   The display device according to any one of claims 1 to 5, wherein the sound acquisition range in the wide angle mode is 60 degrees or more. When the state of the display device changes,
The first acquisition of the audio data is performed in the wide angle mode,
2. The method according to claim 1, wherein a direction of the voice acquisition at the time of acquiring the voice data for the second time or later is set to a determined speaker direction in the first acquisition of the voice data, and the direction is obtained in the narrow angle mode. The display device according to claim 6.   A television receiver comprising: the television receiver according to claim 1.

Images