JP2000338986A - Voice input device, control method therefor and storage medium - Google Patents

Abstract

PROBLEM TO BE SOLVED: To wipe away a resistance sense of speaking to an apparatus thereby freely performing voice input, and also distinguish a voice vocalized to the apparatus from a voice vocalized to normal others. SOLUTION: In this voice input device, a level comparison part 106 in a voice input part 1 compares an inputted voice signal with a threshold value for deciding whether the voice has whisper volume or not, and passes only a voice below the threshold. Thereby, a resistance sense to speaking to an apparatus is wiped away to freely perform voice input, while a voice vocalized to the apparatus can be distinguished from a voice vocalized to normal others.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a voice input device, a control method thereof, and a storage medium.

[0002]

2. Description of the Related Art As a device for controlling a device by voice input, there is an answering machine for inputting an answering machine message or a message by voice, and in recent years, a car navigation system with a voice recognition function and the like. It is also used as a character input means instead of a keyboard. However, there is a situation in which when using such a voice input device, the user does not want to hear the input voice when inputting the voice, or does not want to bother others. Or,
It is difficult to speak loudly in public.

[0003] Japanese Patent Laid-Open No. 5-265472 discloses a speech information processing apparatus which realizes input of speech without being heard by other people around and without being affected by ambient noise. And Japanese Patent Application Laid-Open No. Hei 5-114880.

A speech information processing apparatus disclosed in Japanese Patent Application Laid-Open No. 5-265472 uses a speech reflector having an ellipsoidal structure, and a handset (speaker and microphone) for speech input / output at one focal point of the ellipsoid. Is designed so that the operator's head, especially the ears and the mouth, comes to the other focal point, so that the operator of the device does not have a handset and can input / output information to others who are lined up behind. Can input and output audio information without being heard.

A portable mobile radio terminal disclosed in Japanese Patent Application Laid-Open No. Hei 5-114880 is intended to provide a small-sized mobile radio terminal which does not require a user to make a loud voice even in a place where background noise is high. Then, the speaker's voice is input to a voice recognition device having a plurality of voice dictionaries registered in advance according to the assumed use situation of the speaker, and the speech-recognized word output is changed according to the tone desired by the speaker for conversion. A plurality of speech dictionaries (I
I) is input to the accompanying speech synthesizer, and the synthesized speech is output. As a result, it is possible to extract only syllables that can be recognized as registered speaker's voice from the microphone input. It becomes possible to make a call without issuing a voice, and since voices not registered in the voice dictionary are not recognized, unauthorized use by anyone other than the owner can be substantially prohibited. Furthermore, a synonymous sentence conversion function can be added between speech recognition and speech synthesis, and tone conversion can be performed during speech synthesis, so that the impression given to the other party can be controlled, so that conversation and negotiation can be carried out advantageously.

[0006]

In any case, although the effect of the surrounding noise is reduced and the voice information is surely input to the device even with a small voice, it is necessary to talk to the device. It does not focus on making it easy to perform voice input by eliminating the resistance to voice, and distinguishing between voices uttered to ordinary people and voices uttered to devices is not There was nothing to realize.

SUMMARY OF THE INVENTION The present invention has been made in view of the above-described problems, and has been made in consideration of the above circumstances. It is an object of the present invention to provide a voice input device, a control method thereof, and a storage medium capable of identifying a voice uttered to a device.

[0008]

In order to solve such a problem, for example, a voice input device according to the present invention has the following configuration. That is.

A threshold for determining whether or not the voice signal input from the voice input means is the volume of whispering voice; a voice signal exceeding the threshold is not processed; Is processed and output.

Further, according to a preferred embodiment of the present invention, it is preferable that the apparatus further comprises a voice signal detecting means for detecting whether or not the voice signal is a voice for the voice input symbol. As a result, even when the sound volume is low, noise around non-speech can be excluded.

Preferably, the audio signal detecting means detects whether the audio signal is an audio signal based on a power spectrum and a pitch frequency. This makes it possible to distinguish between the mechanical sound and the voice uttered by a human.

[0012] Further, speech is recognized from the speech signal,
It is desirable to include a voice recognition unit that obtains voice data, and an output unit that outputs information based on the voice data obtained by the voice recognition unit.

In this case, if the output means outputs the character code obtained by voice recognition to the Kana-Kanji conversion process, the output means can function as an input means replacing a conventional keyboard or the like. Will be possible.

If the output means outputs the character code obtained by voice recognition as a command to the conversion processing, it can be made to function as a scanning means instead of a button or a switch.

Further, it is desirable that the voice recognition means has a dictionary pattern when whispering is uttered, and performs voice recognition processing using the dictionary pattern. As a result, by using a dictionary dedicated to whispering voices, it is possible to improve the recognition accuracy of only whispering voices.

[0016]

Embodiments of the present invention will be described below in detail with reference to the accompanying drawings.

<Structure of Speech Input Device> FIG. 1 shows a block diagram of a speech input device of the present embodiment.

In FIG. 1, reference numeral 1 denotes a voice input device according to the present embodiment, which has the following configuration.

Reference numeral 101 denotes a CPU that controls the entire system based on a program stored in a ROM 103 described later. 102
Is a RAM for storing whispering voice dictionary information and other various control data and user data, and 103 is a voice input device 1
Is a ROM for storing the control program.

Reference numeral 104 denotes a voice input unit, which is a high-sensitivity microphone 10 that performs acousto-electric conversion with high sensitivity even for low-volume voice.
5. Comparing the input signal with a predetermined lower limit level, and comparing the input signal with a predetermined upper limit level, a signal amplifying unit 107 for amplifying the signal, calculating a pitch frequency and a power spectrum from the input signal, and It is composed of a signal analysis unit 108 that performs detection.

Reference numeral 109 denotes an RA based on the control of the CPU 101.
A voice recognition unit 110 for performing voice recognition with reference to the whisper voice dictionary information in M102, and a display unit 110 for notifying the operator of the result of voice recognition and the like. Reference numeral 111 denotes a voice synthesis unit that generates a synthesized voice for notifying the operator of the result of voice recognition and the like, 112 denotes a signal amplification unit that amplifies the output synthesized voice of the voice synthesis unit to an appropriate signal level, and 113 denotes a voice recognition unit. This is a speaker (earphone) that outputs a synthesized voice for notifying the operator of a result or the like. A communication unit 114 performs wireless communication with an external device (for example, a personal computer or the like). A device for communicating with the communication unit 114 is also connected to the external device side, through which text is input.

FIG. 2 shows a use situation which is one mode of the voice input device in the embodiment. As shown in the figure, the operator is equipped with a microphone 150 and a speaker 113.

Next, the voice input operation of this embodiment will be described with reference to the flowchart of FIG.

In this embodiment, the RAM 10
It is assumed that training has been performed to adapt the whisper dictionary information in 2 to the whisper of the repeater.

First, in step S101, the level comparator 1 detects a signal input from the high-sensitivity microphone 105.
At 06, a comparison is made with a predetermined lower limit level. Step S
If the result of the comparison in step 101 is less than the predetermined lower limit level, the process returns to step S101. If the result is greater than the predetermined lower limit level, the process proceeds to step S102.

In step S102, in order to capture only a low-volume audio signal, the level comparison section 106 compares the signal with a predetermined upper limit level. As a result of the comparison, if it is larger than the predetermined upper limit level, the process returns to step S101, and if it is smaller than the predetermined upper limit level, the process proceeds to step S103.

As a result, it is possible to pass only a voice lower than the volume of a normal conversation and a voice higher than the volume for removing ambient noise.

In step S103, the signal amplifying unit 1
At 08, the input signal is amplified to an appropriate input level.
Then, in step S104, the signal analyzer 109
The input signal is analyzed in. For example, a pitch frequency and a power spectrum are calculated. Then, step S
If it is determined in step 105 that the signal is not an audio signal based on the analysis result in step S104, the process returns to step S101.
If it can be determined that the signal is an audio signal, the process proceeds to step S106.

In step S106, the voice recognition unit 112
Then, the voice recognition of the input signal is performed with reference to the whisper dictionary information in the RAM 102. Then, step S10
In step 7, based on the voice recognition result in step S106, the operator is notified of the reaction of the voice input. For example, voice information required in response to voice input from the voice synthesis unit 112 is output, a predetermined sound for notifying that recognition has been normally completed is reproduced by a speaker (earphone), or the display unit 111 To display information as a reaction.

In step S108, the recognition result (character code group) is sent to the external device via the communication unit 114.
Notify.

The voice input device of the present embodiment can function as input means instead of a normal keyboard. In addition, it can also function as command input means instead of switches and buttons.

When the external device is, for example, a personal computer, it receives information (a group of character codes indicating kana) from the communication unit 114 and converts it into a kana-kanji conversion program (called FEP or the like) or a command interpretation program. Device driver to be delivered (a type of program)
It should just be made to operate as.

When the present device is a portable device, wireless communication is desirable as the communication unit 114, but it goes without saying that the communication unit 114 may be connected by a physical cable.

In the embodiment, an example has been described in which voice recognition is performed and characters are input instead of a keyboard, or commands are input instead of switches. However, the invention has wide application and is not intended to limit the invention.

Further, a sound source board is mounted as a standard in the information processing apparatus of a personal computer these days,
There is a microphone connection. A microphone may be connected to the microphone connection terminal, and the sound source board, a CPU in a personal computer, and a program for realizing the processing of FIG. 3 may be realized. Alternatively, a personal computer may be provided with a dedicated expansion board for connecting a microphone having good sensitivity.

In this case, the program operates as a device driver or a part of an application program. The difference from FIG.
The only difference is the output target at 108.

As described above, in the present embodiment, in the audio input device, only the audio signal of a predetermined volume or less is input, and it is detected that the input signal is an audio signal. An audio signal detection unit, a configuration in which the audio signal detection unit detects an audio signal based on a power spectrum and a pitch frequency, a voice recognition unit that performs a voice recognition process on the input signal, and a whisper to the voice recognition unit By providing a dictionary pattern in the case of doing and performing a voice recognition process using the dictionary pattern, it is possible to easily perform voice input by eliminating the resistance to talking to the device, There is a great effect of providing a voice input device and method capable of distinguishing a voice uttered to a normal person from a voice uttered to a device. .

In the present embodiment, the form of the high-sensitivity microphone and the main body of the voice input device is different from that shown in FIG. It goes without saying that other forms such as forms may be used.

Steps S101 and S101 in FIG.
The threshold value to be compared at 02 can be appropriately adjusted by an operator (operate a volume knob (not shown)). As a result, it becomes possible to absorb the problem of individual differences in whispering voices or the situation of surrounding noise.

The present invention may be applied to a system composed of a plurality of devices, or may be applied to a device composed of one device (for example, an information processing device with a built-in microphone).

Another object of the present invention is to supply a storage medium (or a recording medium) in which a program code of software for realizing the functions of the above-described embodiments is recorded to a system or an apparatus, and 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 the memory provided in the function expansion card inserted into the computer or the 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.

[0043]

As described above, according to the present invention, it is possible to easily perform voice input by eliminating the resistance to talking to the device, and to make the voice uttered to the other person normal. The voice uttered to the device can be identified.

[Brief description of the drawings]

FIG. 1 is a block diagram of a voice input device.

FIG. 2 is a schematic view of a voice input device.

FIG. 3 is a flowchart of a voice input operation.

[Explanation of symbols]

 Reference Signs List 1 voice input device 101 CPU 102 RAM 103 ROM 104 voice input unit 105 high sensitivity microphone 106 level comparison unit 107 signal amplification unit 108 signal analysis unit 109 voice recognition unit

Claims (8)

[Claims] An audio signal input from an audio input means has a threshold value for determining whether or not the volume of the whispered audio, and an audio signal exceeding the threshold value is set as a non-processing target.
An audio input device that processes and outputs an audio signal equal to or less than the threshold. 2. The apparatus according to claim 1, further comprising an audio signal detecting unit for detecting whether or not the input signal is a voice for the voice input symbol, and processing the audio signal determined to be a voice by the audio signal detecting unit. The voice input device according to claim 1, wherein the voice input device is a target. 3. The voice input device according to claim 2, wherein said voice signal detection means detects whether or not the voice signal is a voice signal based on a power spectrum and a pitch frequency. 4. The apparatus according to claim 1, further comprising: voice recognition means for recognizing voice from the voice signal to obtain voice data; and output means for outputting information based on the voice data obtained by the voice recognition means. The voice input device according to any one of claims 1 to 3. 5. The voice input device according to claim 4, wherein said output means outputs a character code obtained by voice recognition to a kana-kanji conversion process. 6. The voice input device according to claim 4, wherein said voice recognition means has a dictionary pattern when uttered with a whisper, and performs voice recognition processing using said dictionary pattern. 7. A comparison is made between a threshold value for determining whether or not the audio signal has a whisper volume and an actual audio signal input from the audio input means, and an audio signal exceeding the threshold value is not processed. And processing and outputting an audio signal equal to or less than the threshold value. 8. A storage medium for storing a program code which, when read and executed by a computer, functions as a device for processing an audio signal input from the audio input means, wherein the audio signal is input from the audio input means. Has threshold data for determining whether or not the volume of the whispering voice,
A storage medium for storing a program code for processing an audio signal having a threshold value or less and outputting the audio signal having the threshold value or less.