JPS61150030A - Voice input and output device - Google Patents

Abstract

PURPOSE:To attain ease of confirmation of a commanded content by providing plural microphones inputting voice to an input voice recognizing section respectively. CONSTITUTION:A voice output control part 2 receives a voice output command received from a host system 1, uses a voice information storage part to read voice information and outputs it to a voice synthesizer 4. The voice uttered by the voice synthesizer 4 uses a voice output adjusting part 5 to adjust its level and is outputted from a speaker 6. The outputted voice is fetched by plural microphones 14a-14n provided in advance while expecting the moving range of a listener and transmitted to preprocessing parts 11a-11n. The preprocessing parts 11a-11n arrange the received input voice and transmit it to corresponding voice recognizing parts 8a-8n. The voice recognizing parts 8a-8n compare the inputted voice with voice registration information registered in advance, the coincidence of the content is discriminated and the result of discrimination is transmitted to a discrimination part 7. The result of discrimination of the discrimination part 7 is transmitted to the host system 1 via a voice recognition control part 18.

Description

DETAILED DESCRIPTION OF THE INVENTION [Technical Field of the Invention] The present invention relates to a speech input/output device including an input speech recognition processing section and a speech synthesis processing section.

[Technical background of the invention and its problems]

The audio input/output device of the monitoring and alarm system that has an audio output function (i.e., the recognition of various instructions and notifications to the worker (supervisor) when outputting audio, i.e., the attention of (a) was distracted) (b) The output level is inappropriate due to surrounding noise; (C) The audio output level is inappropriate because the speaker is far away. It is necessary to confirm the recognition of the information output to the person.

Conventionally, such confirmation signals and notification means have been either simply by repeatedly outputting the same content several times, or by re-outputting the audio by requesting the worker to re-output the audio. Furthermore, visual communication was used, but such visual communication was easy to miss and was not necessarily easy to use.

[Purpose of the invention]

The present invention aims to solve the above-mentioned problems and to effectively use existing audio input/output devices.

[Summary of the invention]

The present invention includes an input speech recognition processing section and a speech synthesis processing section, and the host data determines whether or not the human voice recognition section can recognize the speech output from the speech synthesis processing section based on a command from the host data processing device. A processing device (an audio input/output device that is configured to output a voice from a voice synthesis processing section, adjust the voice output level in response to a re-voice command from the host data processing device, and output the voice from the voice synthesis processing section) , the input voice recognition unit is equipped with a plurality of microphones for inputting voice, respectively, and is configured so that the content of the output voice can be reliably conveyed according to instructions from the host data processing device.

That is, the present invention includes an input speech recognition processing section and a speech synthesis processing section, and sends the speech recognition processing result by the input speech recognition processing section to a host data processing device, and also outputs an output command from the host data processing device. The audio input/output device configured to output audio from the audio synthesis processing unit is equipped with an audio output adjustment unit that can adjust the level of the audio output to the audio synthesis processing unit, and It is equipped with multiple recognition units for recognizing human voices, and conveys information to the host system as to whether or not the output voice has been reliably recognized.
This voice input/output device is characterized in that, if the voice is not recognized, the voice output level is adjusted according to a command from the host system, and the voice is output from the voice synthesis section.

Regarding the confirmation of voice recognition, in addition to the judgment in the voice recognition processing unit that recognizes the output voice and human voice, the output voice can be confirmed by operating an external push button switch. It can also be an audio input/output device that notifies the host system that it has been recognized.

[Embodiments of the invention]

Next, examples of the present invention will be described. The drawing shows an audio input/output device consisting of the following components.

(b) Audio output control unit 2 that outputs audio information retrieval commands in response to audio output commands received from the host system 1 (0) Audio information storage that outputs audio information corresponding to audio information retrieval commands from the audio output control unit 2 Part 3
, (c) a speech synthesizer 4 that synthesizes the synthesis based on the speech information output from the speech information storage section 3; Audio output adjustment unit 5 that outputs to the speaker 6 (E) Microphone 14a that listens to the audio output from the speaker 6 (F) Preprocessing unit 11a that preprocesses the input audio heard by the microphone 14a (G) Preprocessing unit 11a A voice recognition unit 8a performs voice recognition of the input voice organized in
Judgment unit 7 transmits the judgment result to speech recognition control unit 18 Next, the operation of the embodiment is as follows.

When performing audio output, the audio input/output device transmits the output result to the host system 1 by performing the following steps.

(1) When the audio output control unit 2 receives the audio output command including coded information sent from the host system 1, it retrieves necessary audio output information from the audio information storage unit 3.

(b) The audio output control unit 2 sends the collected audio output information to the audio synthesizer 4 (=.

(C) The speech synthesizer 4 synthesizes speech according to the given speech output information, and the speech output adjustment section 5 outputs the sound to the speaker 6 according to the predetermined sound level.
Perform more audio output.

(d) A plurality of microphones 14a to 1 are provided in advance, assuming the operating range of the person who should listen to the output audio.
Preprocessing unit 11a captures each audio output from 4n.
~ Send to 11r+. The preprocessing units 11a to 11n organize the received input voices and send them to the corresponding voice recognition units 8a to 8n.

(e) The host system ↓ also outputs the same information as the voice output command including coded information sent to the voice output control unit 2 to the voice recognition control unit 18 at the same timing as outputting the command.

(f) As mentioned above, the voice recognition units 8a to 8n extract information from the voice registration information 17 based on the content instructed by the voice recognition control unit 18 and send it to the determination unit 7;
In order to judge whether or not the contents recognized by the speech recognition sections 82 to 8n match, 1: The recognition result is sent to the judgment section 7.

(g) The determination unit 7 compares the contents recognized by the voice recognition units 8a-pr8n from the retrieved voice registration information 17, and determines whether or not the recognition is correct.

(h) The judgment unit 7 transmits the judgment result to the voice recognition control unit 18.

(i) When the voice recognition control unit 18 determines that the voice has been correctly output and recognized, it transmits a voice output level signal to the host system 1.

(j) On the other hand, if it is determined that the recognition is not good, the voice recognition control unit 18 requests the host system 1 to re-output the voice output.

(k) The host system 1 that received the request in item (j) above determines that the audio output level is insufficient and that the listener cannot correctly recognize it, and requests the audio output adjustment unit 5 to adjust the audio output level. to a predetermined level,
At the same time, the audio output level control information 20 is outputted so as to increase by a certain amount (also at a predetermined rate), and at the same time, the above-mentioned (1)
Following the same format as in Section 1, instruct the re-output of the audio output.

(From 11 onwards, repeat this operation in sequence until it is correctly recognized.

(E) On the other hand, when the audio output is performed, if the listener is able to recognize the output audio, the listener presses the push button switch 21 or verbally utters that the audio has been recognized.

(n) When the push button switch 21 is pressed, the judgment unit 7 immediately informs (=speech recognition control unit 18:: that it has been correctly recognized.

(p) If the listener utters something that they recognize, it means that
Preprocessing units 11a to 1 via microphones 14a to 14n
1n to the speech recognition sections 8a to 8n, and the recognition results are sent to the judgment section 7.

(q) The determination unit 7 refers to the voice registration information 17,
Inform the voice recognition control unit 18 that the output voice has been recognized,
The voice recognition control unit 18 provides the host system 1 with the above (i)
), it also conveys the audio output level signal.

(r) The host system 1 completes the audio output work when it receives the signal of the audio output level mentioned above, and waits until the next audio output is required.

〔Effect of the invention〕

According to the present invention, the following effects can be obtained.

(,) When the worker is within the specified action range with voice output:
: If you are there, you can definitely hear it.

(b) Even when there is noise, the output content can be accurately conveyed because the output is repeatedly output while adjusting the audio output level until it can be recognized by the microphone.

(C) Since workers can check the instructions (audio output content) anywhere within their range of action, they can continue working efficiently, and improvements in the man-machine interface can be expected.

[Brief explanation of drawings]

The drawing is a configuration diagram of four audio input/output positions showing an embodiment of the present invention. DESCRIPTION OF SYMBOLS 1... Host system, 2... Audio output control section 3... Audio information storage section, 4... Speech synthesizer 5.
...Audio output adjustment section, 6...Speaker 7...Judgment section, 8a to 8n...Speech recognition section 11
a to 11n... pre-processing section. 14a-14n...Microphones. 17... Voice registration information, 18... Voice recognition control unit 19... Voice input/output device. 20...Audio output level control information. 21...Push button switch

Claims (1)

[Claims] An audio input/output device comprising the following components. (b) Audio output control unit (
2) (b) An audio information storage unit (3) that outputs audio information corresponding to the audio information retrieval command from this audio output control unit (2). A speech synthesizer (4) that synthesizes sound using the speech information obtained (d) The synthesized speech synthesized by this speech synthesizer is adjusted according to a preset speech output level, and then output to the speaker (6).
) (e) A microphone (14a) for listening to the sound output from the speaker (6) (f) Preprocessing for preprocessing the input sound heard by this microphone (14a) Part (11a) (G) Speech recognition section (8a) that performs speech recognition of the input speech sorted by this preprocessing section (11a) (H) Input speech recognized by this speech recognition section (8a) and pre-registered a determining unit (7) that compares the registered voice information, determines whether the contents match, and transmits the determination result to the voice recognition control unit (18);