JP2000284794A - Device and method for voice recognition and voice recognition system - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a device and a method for voice recognition and a voice recognizing system an analog command inputting is made possible employing a voice input only. SOLUTION: If the recognition result of input voice indicates a bearing in a voice recognition device 101, a voice parameter obtaining section 108 obtains elements such as length, volume and interval of the input voice other than vocabulary contents as parameters and an analog amount corresponding to the parameters is determined. Then, in an extended voice command generating section 110, the analog amount is added to a voice command and an extended voice command is generated. In a camera section 112, vocabulary contents of the inputted expanded voice command and its analog amount are extracted and an imaging section 115 is controlled by the portion of the analog amount based on the command.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a voice recognition apparatus and method, and a voice recognition system, and more particularly to a voice recognition apparatus and method capable of inputting a command by voice, and a voice recognition system.

[0002]

2. Description of the Related Art Generally, when it is desired to perform analog input to a device or a system, the input should be performed by a device such as a dial, a bar, or a mouse which assists analog input. However, there are cases where manual input by the user cannot be expected, for example, it is difficult to perform manual input because the device to be input is extremely small.

In order to solve such a problem, the apparatus is provided with a voice input device such as a microphone, so that a voice uttered by a user is input and recognized, and the recognition result is utilized as a command. Devices are known.

[0004]

However, when a command input to the conventional voice recognition device is realized by voice input, it is difficult to perform a delicate analog operation only by the voice input.

[0005] For example, consider a case where a left and right direction instruction command is input to an imaging device capable of voice recognition. In this case, analog operations can be performed by adding an adverb to the command itself, such as "slightly right" or "largely left", for the basic commands "right" and "left". It is also possible.

However, according to this method, only a change amount corresponding to a pre-registered adverb can be controlled. Also, the user cannot determine which adverb can be input as a voice command.

SUMMARY OF THE INVENTION The present invention has been made to solve the above-described problems, and has a voice recognition apparatus and method which enables analog command input by voice input only.
And a speech recognition system.

Also, taking into account the characteristics of the utterance for each user,
It is an object of the present invention to provide a voice recognition device and method and a voice recognition system that enable analog voice command input.

[0009]

As one means for achieving the above object, the speech recognition apparatus of the present invention has the following arrangement.

That is, input means for inputting a voice signal, voice recognition means for recognizing the vocabulary content indicated by the voice signal,
It is characterized by having feature amount obtaining means for obtaining a feature amount in the audio signal, and command creating means for creating a command based on the vocabulary content and the feature amount of the audio signal.

For example, the command creating means creates a command so that the degree of operation indicated by the vocabulary content is indicated by the feature amount.

For example, the characteristic amount indicates an analog change amount in the audio signal.

[0013]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment according to the present invention will be described below in detail with reference to the drawings.

FIG. 1 is a block diagram showing a configuration of a camera system according to the present embodiment. In the camera system according to the present embodiment, the pan / tilt of the camera can be controlled by voice input from a user.

In FIG. 1, reference numeral 101 denotes a voice recognition device which is a feature of the present embodiment, and reference numeral 102 denotes a camera unit which performs actual photographing in accordance with a voice command recognized by the voice recognition device 101. The voice input to the voice recognition device 101 is performed by inputting a voice waveform from a voice input device such as the microphone 104. Alternatively, voice may be input from the telephone 102, the portable terminal 103, or the like via the general public line 105.

In the speech recognition apparatus 101, 10 is a CP
The control unit includes a U10a, a ROM 10b, a RAM 10c, and the like, and controls the operation and processing of each component described later. In particular, the CPU 10a performs control of multiplexing and related processes to be described later using the RAM 10c as a work memory according to a program stored in the ROM 10b in advance.

The voice input to the voice recognition device 101 is
First, the voice is captured by the voice capturing unit 106, and the voice waveform is sent to the voice recognition unit 107 and the voice parameter acquiring unit 108.

The voice recognition unit 107 recognizes the input waveform by voice and determines which voice command the input voice corresponds to. The recognition result determined here is sent to the extended voice command creation unit 110 as a voice command, and is also sent to the voice parameter acquisition unit 108.

The speech parameter acquiring unit 108 acquires, as a parameter, an element (analog element) other than lexical contents, such as a speech length, a volume and a pitch, when the recognition result of the speech recognizing unit 107 indicates an azimuth.
The parameter obtained here is used as the analog amount determination unit 109.
To determine the analog amount that matches the parameter. The determined analog amount is sent to the extended voice command creation unit 110.

The extended voice command creating section 110 adds an analog amount corresponding to each voice parameter determined by the analog amount determining section 109 to the voice command which is the recognition result of the voice recognition section 107.
Hereinafter, the voice command to which the analog amount is added is referred to as an extended voice command. The extended voice command is sent to the extended voice command output unit 111, and the voice recognition device 101
Is output to the camera unit 112 as the recognition result in.

The extended voice command input unit 113 of the camera unit 112 includes general voice commands such as "exposure", "copy", "shutter", "high, cheese", and the like.
Commands indicating directions such as “right” and “up” are input as extended voice commands. These extended voice commands indicating the azimuth are sent to the azimuth parameter control unit 114, where they are decomposed into lexical contents of the command and parameters of the analog amount. The lexical content indicates the operation content, and the analog amount indicates the degree of the operation.

These parameters are sent to the imaging unit 115, and by controlling an actuator (not shown),
The imaging unit 115 changes the pan / tilt angle, for example, by the input analog amount.

According to the present embodiment, namely, the speech recognition device 1
For example, when a voice command indicating a direction of “right” is input to the camera unit 11, if the voice is longer, louder, or treble, for example,
It is possible to operate the second imaging unit 115 to move it to the right.

FIG. 2 is a flowchart showing the operation of the speech recognition apparatus 101 in the present embodiment. The control program for this processing is stored in, for example, the ROM 10b, and is read out and executed on the RAM 10c by the CPU 10a. Hereinafter, the extended voice command output processing according to the present embodiment will be described in detail with reference to FIG.

First, when the voice recognition process is executed, the process enters a voice detection loop (S201).
The voice recognition unit 107 performs voice recognition using the voice waveform (S202). Here, it is conceivable that the result of voice recognition is not output because the uttered content is ambiguous or the like. Therefore, it is determined whether or not a voice recognition result is obtained from the voice recognition unit 107 (S203). If no recognition result is obtained, the recognition process is interrupted.

When the speech recognition result is output, the speech parameter acquiring unit 108 determines whether the recognition result is a word indicating a direction (S204). If it is a word indicating an azimuth, voice parameters are obtained from the input voice waveform (S205). Here, analog elements such as the utterance time (or utterance speed), volume, and pitch information of the voice are acquired as voice parameters. Next, the analog amount determining unit 109 determines the analog amount of the voice command based on the values of the voice parameters (S206).

As a method of determining the analog amount, it is conceivable to prepare a table, for example, as shown in FIG. According to the example shown in FIG. 4, the analog amount is determined according to the state of each voice parameter such as the utterance time, the volume, and the pitch. Of course, this table is not limited to the example shown in FIG. 4. For example, an analog amount may be set by a combination of parameters, and the set analog amount is not limited to two types, large and small. , It is possible to have multiple stages.

As another method of determining the analog amount, it is possible to directly calculate the analog amount from the voice parameter by the following arithmetic expression.

Analog amount = α × voice parameter + b
(Α and b are constants) As described above, the analog amount obtained in step S206 is
The extended voice command is created by being combined with the result of the voice recognition in the extended voice command creating unit 110 (S207). This extended voice command is output to the external camera unit 11 via the extended voice command output unit 111.
2 (S208).

On the other hand, if the speech recognition result is not a word indicating an orientation in step S204, the extended speech command creation unit 110 creates a speech command based on only the recognition result (S209) and outputs it to the camera unit 112. (S210).

FIG. 3 is a flowchart showing the operation of the camera unit 112 corresponding to the extended voice command created as described above. When an extended voice command or a voice command output from the voice recognition device 101 is input to the camera unit 112 according to the flowchart of FIG. 2 described above, the camera unit 112 exits from a voice command waiting loop (S211). Then, it is determined whether the input command is a general voice command or an extended voice command (S212). If it is a general voice command, the control of the imaging unit 115 is performed as it is (S214).

On the other hand, if the command input in step S212 is an extended voice command, the azimuth parameter control unit 114 acquires the azimuth indicated by the command and the analog amount (S213), and then uses these values to perform imaging. The unit 115 is controlled (S214).

As described above, according to the present embodiment,
By detecting an analog element in the utterance of the user and adding an analog amount corresponding to the detected element to the command, the analog operation can be performed only by voice input.

<Modification 1> The present embodiment can be applied to a system for performing user authentication. In this case, it is necessary to register the default utterance of the user at the time of user registration. Specifically, by analyzing the user's default utterance, the utterance time, volume,
As a format of a voice parameter such as a pitch, for example, RAM1
It may be held at 0c or the like.

The system first authenticates the user when the service starts. Then, in subsequent user operations, an expanded voice command is created by comparing the utterance of the user with a default utterance of the user registered in advance.

In this case, the determination of the analog amount is performed based on the difference between the voice parameter in the current utterance and the default voice parameter. For example, the analog amount is determined according to the following equation.

Analog amount = α × (default voice parameter−current voice parameter) + b (a and b are constants). Thus, the analog amount can be determined in consideration of the utterance characteristics of each user. Therefore, even if the utterance state is different for each user, for example, a user who usually has a high voice or a user who has a high utterance speed, an appropriate extended voice command can be set for each user. It is possible to do.

<Modification 2> Further, even when the present embodiment is applied to a system in which user authentication is not performed, it is possible to input an extended voice command in consideration of the utterance characteristics of each user. It is.

For example, it is sufficient to hold the speech parameters before the user's first utterance, and to determine the analog amount of the extended voice command by comparing with the current utterance.

In this case, the determination of the analog amount is performed based on the difference between the speech parameter in the current utterance and the speech parameter before the first utterance. For example, the analog amount is determined according to the following equation.

Analog amount = α × (speech parameter before one utterance−current speech parameter) + b (α and b are constants). Thus, even if the default utterance for each user is not registered in advance, Similar to 1, the analog amount can be determined in consideration of the utterance characteristics of each user.

[0042]

[Other Embodiments] The present invention can be applied to a system composed of a plurality of devices (for example, a host computer, an interface device, a reader, a printer, etc.), and can be applied to a single device (for example, a copying machine). Machine, facsimile machine, etc.).

An object of the present invention is to supply a storage medium (or a recording medium) in which program codes of software for realizing the functions of the above-described embodiments are recorded to a system or an apparatus, and to provide a computer (a computer) of the system or the apparatus. It is needless to say that the present invention can also be achieved by a CPU or an MPU) reading and executing the program code stored in the storage medium. In this case, the program code itself read from the storage medium implements the functions of the above-described embodiment, and the storage medium storing the program code constitutes the present invention. Also,
When the computer executes the readout program code, not only the functions of the above-described embodiments are realized, but also the operating system (OS) running on the computer based on the instructions of the program code.
It is needless to say that a case in which the functions of the above-described embodiments are implemented by performing part or all of the actual processing.

Further, after the program code read from the storage medium is written into a memory provided in a function expansion card inserted into the computer or a function expansion unit connected to the computer, the program code is read based on the instruction of the program code. Needless to say, the CPU included in the function expansion card or the function expansion unit performs part or all of the actual processing, and the processing realizes the functions of the above-described embodiments.

When the present invention is applied to the storage medium, the storage medium stores program codes corresponding to the above-described flowcharts (shown in FIGS. 2 and 3).

[0046]

As described above, according to the present invention, an analog command can be input only by voice input.

Further, the characteristics of the utterance for each user are taken into consideration.
Analog voice command input becomes possible.

[Brief description of the drawings]

FIG. 1 is a block diagram showing a configuration of a camera system according to an embodiment of the present invention;

FIG. 2 is a flowchart illustrating an extended voice command output process according to the embodiment;

FIG. 3 is a flowchart showing an operation corresponding to an extended voice command according to the embodiment;

FIG. 4 is a diagram illustrating an example of an analog amount determination table according to the embodiment.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 101 Speech recognition apparatus 102 Mobile terminal 103 Telephone 104 Microphone 105 General public line 106 Speech acquisition part 107 Speech recognition part 108 Speech parameter acquisition part 109 Analog quantity determination part 110 Extended speech command creation part 111 Extended speech command output part 112 Camera part 113 Extended voice command input unit 114 Azimuth parameter control unit 115 Imaging unit

Claims (17)

[Claims] An input unit for inputting a voice signal; a voice recognition unit for recognizing vocabulary content indicated by the voice signal; a feature value obtaining unit for obtaining a feature value in the voice signal; a vocabulary content of the voice signal And a command creating means for creating a command based on the feature amount. 2. A speech recognition apparatus according to claim 1, wherein said command creating means creates a command such that a degree of operation indicated by said vocabulary content is indicated by said characteristic amount. 3. The speech recognition apparatus according to claim 2, wherein the feature quantity indicates an analog variation in the speech signal. 4. The speech recognition apparatus according to claim 3, wherein the feature quantity indicates utterance speed information. 5. The speech recognition apparatus according to claim 3, wherein the feature quantity indicates volume information. 6. The speech recognition apparatus according to claim 3, wherein the feature quantity indicates pitch information. 7. The speech recognition apparatus according to claim 3, wherein the feature quantity acquiring unit acquires the feature quantity as a difference from a previously input speech signal. 8. The speech recognition apparatus according to claim 7, wherein said feature quantity acquiring means acquires the feature quantity as a difference from a speech signal input before one utterance. 9. The speech recognition apparatus according to claim 3, wherein said feature quantity acquiring means acquires said feature quantity as a difference from a predetermined speech signal. 10. The voice recognition device according to claim 9, wherein the predetermined voice signal is a default voice signal by a user who has uttered the voice signal input by the input unit. 11. The method according to claim 11, further comprising:
11. The speech recognition device according to claim 10, further comprising a holding unit for holding a plurality of users. 12. The speech recognition according to claim 1, wherein the feature quantity acquiring means acquires the feature quantity when the vocabulary content recognized by the speech recognition means is a predetermined content. apparatus. 13. The speech recognition apparatus according to claim 1, wherein said input means inputs said speech signal via a telephone line. 14. The speech recognition apparatus according to claim 1, wherein said input means is a microphone. 15. An inputting step of inputting a voice signal; a voice recognizing step of recognizing a vocabulary content indicated by the voice signal; a feature obtaining step of obtaining a feature of the voice signal; and a vocabulary content of the voice signal. And a command creation step of creating a command based on the feature amount. 16. A speech recognition system in which a speech recognition device and an operation device are connected, wherein the speech recognition device comprises: a speech input unit for inputting a speech signal; and speech recognition for recognizing vocabulary content indicated by the speech signal. Means, a feature amount obtaining means for obtaining a feature amount in the audio signal, a command creating means for creating a command based on the vocabulary content and the feature amount of the audio signal, and a command output for outputting the command to the operation device Means, the operation device comprises: a command input means for inputting the command; an operation content acquisition means for acquiring operation content and degree information of the operation based on the command; and A voice recognition system comprising: a control unit that controls an operation unit based on operation degree information. 17. A recording medium on which a program code for a speech recognition process is recorded, wherein the program code includes at least a code for an input step of inputting a speech signal, and speech recognition for recognizing vocabulary contents indicated by the speech signal. A code of a process, a code of a feature amount acquiring step of acquiring a feature amount in the audio signal, and a code of a command creating step of creating a command based on the vocabulary content and the feature amount of the audio signal. Recording medium.