JP2014092627A - Voice recognition device, voice recognition method and program for the same - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a voice recognition device capable of suppressing false recognition of a voice of a person who does not utter a voice toward the voice recognition device with higher precision.SOLUTION: The voice recognition device 1 includes: a voice recognition unit 131 for converting voice information which collects a voice in front of the voice recognition device 1 into characters and transmitting them; a person detection unit 110 for detecting a person in front of the voice recognition device 1 and outputting person detection information; a face detection unit 121 for outputting face detection information if image information which images a front side of the voice recognition device 1 includes a face of a person directly watching the voice recognition device 1 in the image; a response possibility determination unit 100 for transmitting voice information according to receipt of the voice information from the voice recognition device 1, the person detection information and the face detection information; a response control unit 101 for determining a response message corresponding to the character information received from the response possibility determination unit 100 by selecting one or a plurality of phrases of a response database 141 which stores a response message for a person phrase by phrase; and a response outputting unit 140 for outputting a voice of the response message.

Description

  The present invention relates to a speech recognition apparatus, a speech recognition method, and a program therefor that accurately detect and respond to utterances to the apparatus.

  As a robot voice recognition technology, a technology for accurately detecting and responding to a utterance to a robot is known. This technique distinguishes whether or not the utterance is directed at the robot.

  An example of this technique is described in Patent Document 1. The technique described in Patent Document 1 senses that a person is around based on person detection information output from a camera, a contact sensor, or a person detection sensor, and extracts face information from image information of the camera. When a face is detected, the utterance is recognized as a speech directed to the robot and responds.

JP 2007-155985 A

  However, the technique described in Patent Document 1 described above includes a sensor and a camera in all directions around 360 degrees around the robot, and reacts to any direction in which people are present. Therefore, the technique described in Patent Document 1 has a problem that it misrecognizes the utterance of a person who has not uttered the robot.

  An object of the present invention is to provide a speech recognition apparatus, speech recognition method, and program that can solve the above-described problems.

A first speech recognition device according to an aspect of the present invention includes a speech input unit that collects speech in front of the speech recognition device and outputs the speech information, and converts the speech information received from the speech input unit into character information. A voice recognition unit that detects a person in front of the voice recognition device and outputs person detection information, a person photographing unit that images the front of the voice recognition device and outputs image information, and the person From the image information received from the imaging unit, a face detection unit that detects the face of a person who is directly looking at the voice recognition device and outputs face detection information; reception of the character information from the voice recognition unit; and In response to reception of the person detection information from the person detection unit and the face detection information from the face detection unit, a response availability determination unit that outputs the character information; and a response message from the voice recognition device to the person. The frame A response database that is databased in units and stored; a response control unit that determines a response message corresponding to the character information received from the response availability determination unit by selecting one or more phrases of the response database; and A response output unit configured to output the response message determined by the response control unit as a voice;

  In a first speech recognition method according to an aspect of the present invention, the speech input unit collects speech in front of the speech recognition device to obtain speech information, and the speech recognition unit uses the speech information received from the speech input unit as characters. In response to the reception of the voice information from the voice recognition unit, the person detection information from the person detection unit, and the reception of face detection information from the face detection unit, the character information is converted into information. The response database stores the response message from the voice recognition device to the person as a database for each phrase, and the response control unit responds to the voice information received from the response availability determination unit. Is determined by selecting one or more phrases in the response database, and the response output unit outputs the voice of the response message.

  In a first program according to an aspect of the present invention, the voice input unit collects voice in front of the voice recognition device to obtain voice information, and the voice recognition unit converts the voice information received from the voice input unit into character information. And the response availability determination unit outputs the character information in response to reception of the voice information from the voice recognition unit and reception of person detection information from the person detection unit and face detection information from the face detection unit. Then, the response database stores the response message from the voice recognition device to the person as a database in a phrase unit, and the response control unit responds to the voice information received from the response availability determination unit, One or more phrases of the response database are selected and determined, and the response output unit causes the computer to execute a process of outputting the voice of the response message.

  According to the present invention, it is possible to prevent erroneous recognition of a voice of a person who is not speaking to the voice recognition device with higher accuracy.

FIG. 1 is a block diagram showing the configuration of the first embodiment of the present invention. FIG. 2 is a diagram illustrating a hardware configuration in which the speech recognition device 1 according to the first embodiment is realized by a computer device and its peripheral devices. FIG. 3 is a diagram illustrating an example of the appearance of the speech recognition apparatus 1 according to the first embodiment. FIG. 4 is a flowchart showing the voice information input operation of the voice recognition apparatus 1 according to the first embodiment. FIG. 5 is a flowchart showing the person detection information input operation of the speech recognition apparatus 1 according to the first embodiment. FIG. 6 is a flowchart showing the face detection information input operation of the speech recognition apparatus 1 according to the first embodiment. FIG. 7 is a flowchart showing a response availability determination operation of the speech recognition apparatus 1 according to the first embodiment. FIG. 8 is a flowchart showing a response message output operation of the speech recognition apparatus 1 according to the first embodiment.

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

[First Embodiment]
FIG. 1 is a block diagram showing a configuration of a speech recognition apparatus 1 according to the first embodiment of the present invention. Referring to FIG. 1, the speech recognition apparatus 1 includes a response availability determination unit 100, a response control unit 101, a person detection unit 110, a person photographing unit 120, a face detection unit 121, a voice input unit 130, and a voice. A recognition unit 131, a response output unit 140, and a response database 141 are provided.

  Next, the configuration of the speech recognition apparatus 1 in the first embodiment will be described.

  The voice input unit 130 collects the voice in front of the voice recognition device 1 and transmits it to the voice recognition unit 131.

  The voice recognition unit 131 includes a voice recognition engine. Specifically, the speech recognition unit 131 discriminates and eliminates noise information such as daily noise that is not related to the utterance of the person from the input speech, and the speech information from which the noise information is excluded by the preprocessing circuit And a speech recognition conversion circuit that recognizes a character corresponding to the uttered sound from the speech information, converts the character to the character information, and outputs the character information to the response availability determination unit 100.

  The person detection unit 110 detects a person in front of the voice recognition device 1 and outputs the person detection information to the response availability determination unit 100. The method for detecting a person may be any method such as a detection method using an infrared sensor or an ultrasonic sensor, or a method for processing and detecting image information of the person photographing unit 120. The person detection information is information indicating or notifying that a person has been detected.

  The person photographing unit 120 photographs the front of the voice recognition device 1 with a camera or the like, and transmits image information to the face detecting unit 121.

  The face detection unit 121 executes processing for detecting a human face from the received image information. As a result, when the face detection unit 121 determines that the face of the person who is directly looking at the speech recognition apparatus 1 can be extracted from the received image, the face detection information is output to the response availability determination unit 100. The face detection information is information indicating or notifying that a face has been detected. The face detection information does not need to be face image information, or does not require information for identifying a face individual.

  When receiving the character information corresponding to the voice information from the voice recognition unit 131, the response availability determination unit 100 checks whether the person detection information and the face detection information are received at the same time. When receiving, the response availability determination unit 100 transmits the character information corresponding to the voice information to the response control unit 101.

  The response database 141 stores a response message for responding to the voice uttered by the person recognized by the person detection information and the face detection information from the voice recognition device 1 in units of phrases and stores the message. The phrase data stored in the response database 141 may be voice data or text information.

  The response control unit 101 selects one or a plurality of phrases in the response database 141 as response messages corresponding to the character information received from the response availability determination unit 100, and outputs the response message to the response output unit 140 as an audio signal. When the phrase data stored in the response database 141 is character information, the response control unit 101 synthesizes the character information of the message of the selected phrase into a voice signal and converts the voice information to the response output unit 140. Output.

  The response output unit 140 outputs the voice of the received response message.

  FIG. 3 is a diagram showing an example of the appearance of the speech recognition apparatus 1 according to the first embodiment of the present invention. As shown in FIG. 3, the speech recognition apparatus 1 is a robot as an example.

  As shown in FIG. 3, the voice input unit 130 of the voice recognition device 1 is mainly composed of a microphone. The person photographing unit 120 is mainly composed of a camera. The response output unit 140 is mainly composed of a speaker. The person detection unit 110 is mainly composed of sensors. In addition, when detecting the person by processing the image information of the person photographing unit 120, the person detecting unit 110 can be omitted. There may be one or a plurality of components, and the installation site is not limited, but the person photographing unit 120 and the person detecting unit 110 photograph and detect the front of the voice recognition device 1.

  FIG. 2 is a diagram showing a hardware configuration in which the speech recognition apparatus 1 according to the first embodiment of the present invention is realized by a computer and its peripheral devices. As shown in FIG. 2, the speech recognition device 1 includes a CPU 11, an output device 12, an input device 13, a main storage device 14, and a secondary storage device 15.

  The CPU 11 operates the operating system to control the entire speech recognition apparatus 1 according to the first embodiment of the present invention. Further, the CPU 11 reads out programs and data from the secondary storage device 15 to the main storage device 14, for example. The CPU 11 functions as the response availability determination unit 100, the response control unit 101, the face detection unit 121, and the voice recognition unit 131 in the first embodiment, and executes various processes based on program control.

  The output device 12 is realized by a speaker or the like, for example, and functions as the response output unit 140 in the first embodiment.

  The input device 13 is realized by, for example, a sensor, a camera, a microphone, and the like, and functions as the person detection unit 110, the person photographing unit 120, and the voice input unit 130 in the first embodiment.

  The secondary storage device 15 is, for example, an optical disk, a flexible disk, a magnetic optical disk, an external hard disk, a semiconductor memory, or the like, and records a computer program so that it can be read by a computer. The computer program may be downloaded from an external computer (not shown) connected to the communication network. The secondary storage device 15 functions as the response database 141 in the first embodiment.

  Note that the block diagram (FIG. 1) used in the description of the first embodiment shows functional unit blocks, not hardware unit configurations. These functional blocks are realized by the hardware configuration shown in FIG. However, the means for realizing each unit included in the speech recognition apparatus 1 is not particularly limited. In other words, the voice recognition device 1 may be realized by one physically coupled device, or by two or more physically separated devices connected by wire or wirelessly, and realized by these plural devices. Also good.

  Further, the CPU 11 reads a computer program recorded in the secondary storage device 15 and operates as a response availability determination unit 100, a response control unit 101, a face detection unit 121, and a voice recognition unit 131 according to the program. Good.

  Further, a recording medium (or storage medium) in which the above-described program code is recorded may be supplied to the speech recognition apparatus 1, and the speech recognition apparatus 1 may read and execute the program code stored in the recording medium. That is, the present invention also includes a secondary storage device 15 that temporarily or non-temporarily stores software (speech recognition program) to be executed by the speech recognition device 1 according to the first embodiment.

  The operation of the speech recognition apparatus 1 configured as described above will be described with reference to the flowcharts of FIGS.

  FIG. 4 is a flowchart showing an outline of voice information input operation in the voice recognition apparatus 1 according to the first embodiment. Note that the processing according to this flowchart may be executed based on the above-described program control by the CPU.

  As shown in FIG. 4, first, the voice input unit 130 collects the voice in front of the voice recognition device 1 and outputs it to the voice recognition unit 131 (step S101).

  Next, the voice recognition unit 131 processes the received voice. Specifically, the preprocessing circuit discriminates and eliminates daily noise that is not related to the voice of the person (step S102).

  Next, the voice recognition unit 131 confirms whether or not a person's utterance is included in the voice information processed by the preprocessing circuit (step S103). If it is present (YES in step S103), the speech recognition conversion circuit of the speech recognition unit 131 converts the speech information into character information, for example, text data, and outputs it to the response availability determination unit 100 (step S104).

  Thus, the voice recognition device 1 ends the voice information input operation.

  FIG. 5 is a flowchart showing an outline of the operation of inputting the person detection information in the speech recognition apparatus 1 according to the first embodiment. Note that the processing according to this flowchart may also be executed based on the above-described program control by the CPU.

  As shown in FIG. 5, first, when detecting a person in front of the voice recognition device 1 (YES in step S201), the person detection unit 110 outputs person detection information to the response availability determination unit 100 (step S202).

  Thus, the voice recognition device 1 ends the operation of inputting the person detection information.

  FIG. 6 is a flowchart showing an outline of the operation of face detection information input in the speech recognition apparatus 1 according to the first embodiment. Note that the processing according to this flowchart may also be executed based on the above-described program control by the CPU.

  As shown in FIG. 6, first, the person photographing unit 120 photographs the front of the voice recognition device 1 and outputs image information to the face detecting unit 121 (step S301).

  Next, the face detection unit 121 performs image processing on the received image information (step S302).

  Next, the face detection unit 121 confirms whether the face of the person who is directly looking at the speech recognition apparatus 1 can be extracted from the image as a result of the image processing (step S303). If it can be extracted (YES in step S303), the face detection information is output to the response availability determination unit 100 (step S304).

  Thus, the voice recognition device 1 ends the operation of inputting the face detection information.

  FIG. 7 is a flowchart illustrating an outline of an operation for determining whether or not response is possible in the speech recognition apparatus 1 according to the first embodiment. Note that the processing according to this flowchart may also be executed based on the above-described program control by the CPU.

  As illustrated in FIG. 7, first, the response availability determination unit 100 receives character information corresponding to collected voice information from the voice recognition unit 131 (step S401).

  Next, the response availability determination unit 100 confirms whether the person detection information from the person detection unit 110 and the face detection information from the face detection unit 121 are received at the same time (step S402). If it has been received (YES in step S402), it is determined that a response is possible, and character information is transmitted to the response control unit 101 (step S403).

  Thus, the voice recognition device 1 ends the response availability determination operation.

  FIG. 8 is a flowchart showing an outline of a response message output operation in the speech recognition apparatus 1 according to the first embodiment. Note that the processing according to this flowchart may also be executed based on the above-described program control by the CPU.

  As shown in FIG. 8, first, the response control unit 101 determines a response message corresponding to the received character information by selecting one or more phrases in the response database 141 (step S501).

  Next, the response control unit 101 transmits the determined response message as a voice signal to the response output unit 140 (step S502). In this case, when the voice message is character information, the response control unit 101 synthesizes the character information, converts it into a voice signal, and outputs the voice signal. If the response message is audio data, the audio data is converted into an audio signal and output to the response output unit 140.

  Next, the response output unit 140 outputs the voice of the received response message (step S503).

  Thus, the voice recognition device 1 ends the response message output operation.

  Next, effects of the first exemplary embodiment of the present invention will be described.

  The voice recognition device 1 in the present embodiment described above can prevent erroneous recognition of the voice of a person who has not spoken to the voice recognition device 1 with higher accuracy.

  This is because the following configuration is included. That is, first, when the person detection unit 110 detects a person in front of the voice recognition device 1, the person detection unit 110 outputs the person detection information to the response availability determination unit 100. Secondly, the person photographing unit 120 photographs the front of the voice recognition device 1. If there is a face of a person looking directly at the speech recognition apparatus 1 in the image, the face detection unit 121 transmits the face detection information to the response availability determination unit 100. Third, the response availability determination unit 100 determines that a response is possible when the person detection information and the face detection information are simultaneously received. As a result, the voice recognition device 1 can respond when a person is in front of the voice recognition device 1 and is directly looking at the voice recognition device 1. An effect is obtained that erroneous recognition of utterance can be prevented with higher accuracy.

  Each component in each embodiment of the present invention described above can be realized by a computer apparatus and firmware based on program control as well as by realizing the function in hardware. The program is provided by being recorded on a computer-readable recording medium such as a magnetic disk or a semiconductor memory, and is read by the computer when the computer is started up. The read program causes the computer to function as a component in each of the embodiments described above by controlling the operation of the computer.

  Although the present invention has been described with reference to each embodiment, the present invention is not limited to the above embodiment. Various changes that can be understood by those skilled in the art can be made to the configuration and details of the present invention within the scope of the present invention.

  For example, each component described in each of the above embodiments does not necessarily have to be individually independent. For example, for each component, a plurality of components may be realized as one module, or one component may be realized as a plurality of modules. Each component is configured such that a component is a part of another component, or a part of a component overlaps a part of another component. Also good.

  Further, in each of the embodiments described above, a plurality of operations are described in order in the form of a flowchart, but the described order does not limit the order in which the plurality of operations are executed. For this reason, when each embodiment is implemented, the order of the plurality of operations can be changed within a range that does not hinder the contents.

  Furthermore, in each embodiment described above, a plurality of operations are not limited to being executed at different timings. For example, another operation may occur during the execution of a certain operation, or the execution timing of a certain operation and another operation may partially or entirely overlap.

  Furthermore, in each of the embodiments described above, a certain operation is described as a trigger for another operation, but the description does not limit all relationships between the certain operation and the other operations. For this reason, when each embodiment is implemented, the relationship between the plurality of operations can be changed within a range that does not hinder the contents. The specific description of each operation of each component does not limit each operation of each component. For this reason, each specific operation | movement of each component may be changed in the range which does not cause trouble with respect to a functional, performance, and other characteristic in implementing each embodiment.

  INDUSTRIAL APPLICABILITY The present invention can be applied to a voice recognition device that can suppress erroneous recognition of a voice of a person who has not spoken to the voice recognition device with higher accuracy.

1 Voice recognition device 11 CPU
DESCRIPTION OF SYMBOLS 12 Output device 13 Input device 14 Main storage device 15 Secondary storage device 100 Response availability determination part 101 Response control part 110 Person detection part 120 Person imaging | photography part 121 Face detection part 130 Voice input part 131 Voice recognition part 140 Response output part 141 Response The database

Claims (6)

A voice input unit that collects the voice in front of the voice recognition device and outputs it as voice information;
A voice recognition unit that converts the voice information received from the voice input unit into character information;
A person detection unit that detects a person in front of the voice recognition device and outputs person detection information;
A person photographing unit that photographs the front of the voice recognition device and outputs image information;
A face detection unit that detects a face of a person who is directly looking at the voice recognition device from the image information received from the person photographing unit, and outputs face detection information;
Response availability determination that outputs the character information in response to reception of the character information from the voice recognition unit and reception of the person detection information from the person detection unit and the face detection information from the face detection unit And
A response database from the voice recognition device to a person, databased in units of phrases, and a response database for storing,
A response message corresponding to the character information received from the response availability determination unit, a response control unit that determines by selecting one or more phrases of the response database; and
And a response output unit that outputs the response message determined by the response control unit as a voice.   The speech recognition apparatus according to claim 1, wherein the speech recognition unit includes a preprocessing circuit that discriminates and eliminates daily noises that are not related to a person's utterance from the speech information.   The speech recognition apparatus according to claim 1, wherein the face detection unit performs image processing on the image information received from the person photographing unit for face detection.   The speech recognition apparatus according to claim 1, wherein the response message stored in the response database is speech data or character information. The voice input unit collects the voice in front of the voice recognition device as voice information,
A voice recognition unit converts the voice information received from the voice input unit into character information;
A response availability determination unit that outputs the character information in response to reception of the voice information from the voice recognition unit and reception of person detection information from the person detection unit and face detection information from the face detection unit;
The response database stores the response message from the voice recognition device to the person as a database in phrase units,
The response control unit determines a response message corresponding to the voice information received from the response availability determination unit by selecting one or more phrases in the response database,
A voice recognition method in which a response output unit outputs voice of the response message. The voice input unit collects the voice in front of the voice recognition device as voice information,
A voice recognition unit converts the voice information received from the voice input unit into character information;
A response availability determination unit that outputs the character information in response to reception of the voice information from the voice recognition unit and reception of person detection information from the person detection unit and face detection information from the face detection unit;
The response database stores the response message from the voice recognition device to the person as a database in phrase units,
The response control unit determines a response message corresponding to the voice information received from the response availability determination unit by selecting one or more phrases in the response database,
A program that causes a computer to execute a process in which a response output unit outputs the voice of the response message.

Images