JPS59228434A - Voice input control system - Google Patents

Abstract

PURPOSE:To control freely a voice recognizing device at a fixed atation side from a mobile station side by controlling the changeover of recognizable or unrecognizable state by means of a control signal having a desired frequency and length. CONSTITUTION:A voice input control section 4 detects whether or not a voice signal received by a fixed station 2 is a control signal having a prescribed frequency and a prescribed time or over, and every time the control signal is detected, a voice input changeover section 3 is controlled to change over the voice input state where the voice signal received by the fixed station 2 is inputted to the voice recognizing section and the voice non-input state where the voice signal is not inputted to the voice recognizing section 5 to control a voice output section, thereby outputting a message attended with the changeover.

Description

[Detailed Description of the Invention] [Technical Field of the Invention] The present invention controls switching between a recognizable state and an unrecognizable state of a voice input/output device having a voice recognition section using a control signal of a predetermined frequency and length. The present invention relates to a voice input control method that enables

[Prior art and problems]

Conventional remote voice input devices are equipped with a voice input device having a voice recognition unit on the fixed station side of the wireless device, input voice from the mobile station side of the wireless device, and perform voice recognition on the fixed station side. Since the voice input device is always in a state where voice recognition is possible, if voice input is performed under high noise conditions from the mobile station side, the fixed station side may only recognize the noise even when voice input is interrupted. , the voice recognition device sometimes operated unnecessarily. In addition, the mobile station may want to exchange private language or temporarily interrupt voice input, but even in such cases, the fixed station remains in a state where it can recognize voice, so it will not be able to exchange private language or electrical data. Conversations were also recognized, making it impossible to temporarily interrupt voice input to make financial transactions.

[Purpose of the invention]

The present invention is based on the above considerations, and provides a remote voice input control system that allows the mobile station to freely (arbitrarily) control the voice recognition unit and disabled state of the voice recognition device on the fixed station side. The purpose is to provide the following.

[Structure of the invention]

To this end, the voice input control method of the present invention includes a voice recognition unit that recognizes an input voice signal, and a switch between the voice recognition unit and the voice recognition unit from a recognition enabled state to an unrecognizable state, or from an unrecognizable state to a recognition enabled state. a voice input/output device having a voice output unit that outputs a message when the message is received; and a fixed station that receives voice signals from a mobile station and transmits a message to the mobile station; and a voice signal that is input to the voice recognition unit. , the audio input/output device is provided with an audio input control section and an audio input switching section, and the audio input control section is configured to control the audio signal received at the fixed station at a predetermined frequency. and detecting whether or not the control signal lasts for a predetermined time or longer, and controlling the audio input switching unit each time a control signal is detected so that the audio signal received at the fixed station is input to the audio recognition unit. The present invention is characterized by being configured to perform switching between an audio input state in which the audio signal is input to the audio recognition unit and an audio unattended state in which the audio signal is not input to the audio recognition unit, and to control the audio output unit to output a message associated with the switching. That is.

[Embodiments of the invention]

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

FIG. 1 is a diagram showing the configuration of one embodiment of the present invention.

In FIG. 1, 1 is a mobile station, 2 is a fixed station, 3 is a filter section, 4 is an audio input control section, 5 is a speech recognition section, 6 is an audio input/output device control section, 7 is an audio output section, and 8 is a Indicates head set. The fixed station 2 transmits and receives voice band signals to and from the mobile station 1, and the voice input signals and control signals received from the mobile station 1 are passed through the filter section 3 and the voice input control section 4.
A message (guidance) in response to the voice input from the voice output section 7 is transmitted to the mobile station 1. The audio input control unit 4 includes a detection circuit 41 and a JK flip-flop 42, and the detection circuit 41 generates a constant frequency signal longer than the audio input time specified by the audio recognition unit, for example, 1 second or more at IKHz. The continuous control signal is detected and the JK rough flip intermediate flop 42 is inverted (on/off) using the output. The filter section 3 includes an analog switch 31 controlled by the output of a JK flip-flop and a filter 32 that analyzes the speaker's voice spectrum. The speech recognition section 5 recognizes the speech of the speaker, and includes a dictionary, a matching circuit, etc., although not shown. The audio input/output device control unit 6 controls the audio input/output device, and is connected to a host computer or the like. The audio output unit 7 outputs guidance and the like in response to audio input, and its output is controlled by the audio input/output device control unit 6. The head set 8 includes a microphone and earphones, and outputs voice input from the speaker and guidance to the speaker.

The mobile station 1 is equipped with a means for generating a control signal, transmits the generated control signal and an input signal inputted by voice from the head set to the fixed station 2, and receives guidance etc. from the fixed station 2. The received guidance etc. are outputted through the earphones of the headset 8 and notified to the speaker.

Next, the operation will be explained. First, when the voice input/output device is in a state where voice recognition is not possible, that is, when the voice unattended terminal ■ is selected by the analog switch 31 in the filter unit 3, even if the speaker inputs voice, the voice recognition unit Since no signal is sent to 5, it is not recognized by the audio input/output device.

Therefore, when the speaker wants to input voice, a means for generating a control signal of the mobile station 1, such as a switch, generates a control signal at a frequency and a predetermined time determined by the specifications of the detection circuit 41 (for example, 1 at IK Hz). (seconds or more) to the fixed station 2. When the fixed station 2 receives the control signal, it sends it to the audio input control section 4 and the filter section 3. In the audio input control unit 4, the detection circuit 41 detects whether an IKHz signal has been input for 1 second or more, and if the signal is valid, one clock pulse is sent to the JK flip-flop 42 at the end of the control signal. send. The clock pulse causes the JK flip flop 42 to invert its output. When the output of the JK flip-flop 42 is reversed, the analog number switch 31 changes from the conventional audio unattended terminal ■ to the audio input terminal ■
After that, the voice input/output device becomes capable of voice recognition. Also, until then, the analog switch 31 has selected the audio unattended terminal in the filter section 3, so the control signal sent to the audio input terminal is not sent to the filter 32, and the control signal is recognized. It will not be done. When the output of the JK flip-flop 42 is inverted, the voice input/output device control section 6 outputs a message output command to the voice output section 7.
Issued to. Upon receiving the message output command, the voice output section 7 notifies the speaker at the mobile station through the fixed station 2 of a message that can be input by voice. Therefore, the speaker can know that the voice input/output device has transitioned to a state in which voice recognition is possible, and can perform voice input. Note that even if the signal is sent to the filter unit 4 during audio input, since the audio input signal does not exist for more than IKHzI seconds, the detection circuit 41 does not detect a valid signal, and the JK flip-flop The output of 42 is also not inverted. therefore,
The analog switch 31 is never switched during voice input.

When the voice input/output device is in a voice recognition enabled state, that is, when the voice input terminal ■ is selected in the analog switch 31 of the filter section 3, the voice signal is sent from the filter section 3 to the voice recognition section 5. Voice recognition processing is performed. Since the input time of the voice is shorter than the input time of the control signal, the detection circuit 41 of the voice input control unit 4 cannot obtain a detection output of the voice signal, and the voice recognition unit 5 Control signals that are too long will be rejected and will not be recognized. When the speaker must have a conversation with a neighbor or a person on the fixed station side (conversation via normal wireless communication) or when high noise is expected to occur, the speaker must shift from the unrecognizable state to the recognizable state. Same as when
When the speaker operates a means for generating a control signal for the mobile station 10, such as a switch, and transmits the control signal from the mobile station 1 to the fixed station 2, the JK flip-flop is inverted in the voice input control section 4, and the analog switch is activated. In step 31, the voice unattended terminal is selected, and thereafter the voice input/output device is in a state in which voice recognition is impossible. At this time, as described above, the voice input/output device control section 4 sends the message output command to the voice output section 7.
Issued to. In response to the message output command, the voice output section 7 notifies the speaker on the mobile station side of the message that voice input is not possible through the fixed station 2.

FIG. 2 is a diagram showing the configuration of an embodiment of a mobile station to which the present invention is applied, in which 11 represents a wireless device, 12 represents an oscillator, and 13 represents a switch. In FIG. 2, when the switch 13 is connected to the microphone side of the headset as shown, the input audio signal from the microphone is sent to the wireless device 11 and transmitted to the fixed station side.
3 is connected to the oscillator 12 side, contrary to the illustration, a constant frequency signal from the oscillator 12 is sent to the wireless device 11 and transmitted to the fixed station side. The oscillator 12 is, for example, one with an oscillation frequency of IKHz as mentioned above.With a simple configuration by switching the switch 12 as described above, it is possible to send a continuous control signal of IKHz for one second or more to the fixed station side. I can do it.

Note that, without being particularly limited to the above description, the audio input/output device of the present invention can be used not only for transmitting and receiving signals between a mobile station and a fixed station wirelessly, but also for transmitting and receiving signals between a mobile station and a fixed station by wire. Furthermore, it goes without saying that not only radio waves but also other materials such as light can be used as the wireless signal propagation medium. In addition, the filter 32 shown in the figure can be included in the voice recognition section 5, or the voice output section 7 can be directly controlled by the output of the voice input control section 4, or its configuration can be partially changed. Needless to say, it is possible.

〔Effect of the invention〕

As is clear from the above description, according to the present invention, the speaker can freely control whether or not the speech recognition unit is enabled.
The recognition rate of the speech recognition device can be increased without recognizing temporary high noise. Furthermore, it is also possible for a person at a fixed station and a speaker at a mobile station to have a wireless conversation, or for a speaker to have a direct conversation with a neighbor.

[Brief explanation of drawings]

FIG. 1 is a diagram showing the configuration of one embodiment of the present invention, and FIG. 2 is a diagram showing the configuration of one embodiment of a mobile station to which the present invention is applied. DESCRIPTION OF SYMBOLS 1... Mobile station, 2... Fixed station, 3... Filter section, 4... Voice input control section, 5... Voice recognition section, 6.
...Audio input/output device control section, 7...Audio output section, 8.
...Head φ set, 11... Radio device, 12...
Oscillator, 13... switch. Patent Applicant: Fujitsu Limited Representative Patent Attorney Yotsube Kyotani

Claims (1)

[Claims] A voice recognition unit that recognizes an input voice signal, and a voice output unit that outputs a message when the voice recognition unit is switched from a recognitionable state to an unrecognizable state or from an unrecognizable state to a recognitionable state. and a fixed station that receives a voice signal from a mobile station and transmits a message to the mobile station, the voice input control method controlling the voice signal input to the voice recognition unit, The audio input/output device is provided with an audio input control section and an audio input switching section, and the audio input control section determines whether the audio signal received at the fixed station is a control signal of a predetermined frequency and a predetermined duration or longer. At the same time, each time a control signal is detected, the audio input switching unit is controlled so that the audio signal received at the fixed station is input to the audio recognition unit. A voice input control system characterized in that the voice input control method is configured to perform a switch between an unmanned state in which voice is not input to a recognition unit, and to control the voice output unit to output a message associated with the change.