JPH0459640B2 - - Google Patents

Abstract

PURPOSE: To prevent the generation of malfunction in important processing due to noise by regulating processing to be executed after voice recognition in accordance with a peripheral noise. CONSTITUTION: A peripheral noise collected by a microphone 6 is detected by a noise level detector 10 and outputted as a DC level corresponding to a noise level. The output level is supplied to discriminators 11a to 11c and compared with reference voltage levels generated from respective resistors 12a to 12c and supplied to the discriminators 11a to 11c. A control part 13 preferentially regulates the processing of an important processing item in accordance with the discrimination results of the discriminators 11a to 11c. Consequently the generation of malfunction in important processing on a noisy place can be prevented.

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Industrial Application] The present invention relates to a telephone with a voice recognition device that recognizes voice and performs processing according to the recognition result.

[Conventional technology]

In recent years, the recognition rate of voice recognition devices has become practical, and they are being used in many places, and devices that are built into telephones and perform processing according to the recognition results have also been proposed. In many cases, when the speaker of this device calls the control destination and instructs the control destination by voice, ``Start heating,'' for example, as in the Arabian Nights, the control destination turns on the heating switch.

[Problem that the invention seeks to solve]

However, even if such a device does not have any problems in an area where there is no ambient noise, it may become inoperable or malfunction in a noisy area, causing problems. Furthermore, if the input voice phrase is too long or too short and falls outside the allowable range for proper recognition, there is also the problem of non-operation or malfunction.

[Means for solving problems]

In order to solve such problems, this invention
A phone with a voice recognition device that recognizes voice and performs processing according to the recognition result includes a noise level detector that detects the ambient noise level and a device that detects which category of a predetermined range the detected noise level belongs to. a discriminator for identifying, a control unit for regulating the above-mentioned processing content according to the discrimination result, and a control unit for determining whether the length of an input voice phrase is within a predetermined allowable range and for determining that the length is outside the range. and means for displaying information corresponding to the purpose.

[Effect]

Processing of important processing items is prioritized and regulated according to the ambient noise level. Furthermore, if the length of the input voice phrase is outside a predetermined allowable range, a display corresponding to that fact is displayed.

〔Example〕

FIG. 1 is a block diagram showing one embodiment of the present invention, in which 1 is a handset, 2 is an automatic level control device, 3 is a voice recognition device, 4 is a CPU, 5 is a display, and 6 is a noise collecting device. Microphone 7 is an amplifier.

FIG. 2 is a functional block diagram showing the essential parts of the present invention in the device shown in FIG. Output as DC level. This output level is determined by the discriminator 11
a to 11c, and is compared with a reference voltage generated by resistors 12a, 12b, 12c, and 12d supplied thereto. In this case, the discriminator 11c has a voltage of V ₁ or higher, and the discriminator 11b has a voltage of V _{2 or more.}
As described above, the discriminator 11a is designed to discriminate _V3 or higher. Reference numeral 13 denotes a control section, which performs the following processing restrictions in accordance with the level detected by the discriminator 1 in accordance with the discrimination results of the discriminators 11a to 11c.

Level 1 (discrimination by discriminator 11a)...Outside call function regulation such as automatic outside line dialing and unlocking, security function regulation Level 2 (discrimination by discriminator 11b)...Hold re-answer, simultaneous call, extension call, etc. extension function regulation, hold function regulation level 3 (discrimination by discriminator 11c)...speaker,
Local function regulations such as clocks, messages, playback, etc. These regulations have the highest level of priority and are regulated first.

FIGS. 3 and 4 are flowcharts when the above operations are processed by the CPU. In the figure, when the data processing in steps 100 to 102 is completed, a change in hook and a change from on-hook to off-hook are detected in steps 103 and 104. When off-hook is detected, one of the states in steps 105 to 107 is recognized, and steps 108, 109, and 11 are performed accordingly.
The combination processing of 0, 111, 112, 113 or the processing of step 114 is performed. and levels 1, 2 of steps 109, 111, 113,
3, in order to determine whether the length of the input voice phrase is within a predetermined recognition tolerance range, a determination is made in step 115, and if the result is ``YES'', step 116 is performed. If the answer is "NO", the judgment in step 117 is made; if the answer is "YES", the message "the sound is too long" is displayed, and the reason for non-recognition is displayed according to each judgment result. do.

If it is determined in step 104 that the change from on-hook to off-hook is "NO", processing for clearing the voice recognition display is performed as shown in step 119. After the determination of "NO" in step 117 and any one of steps 118, 116, 114, and 119 are performed, step 1 is executed.
Twenty other processes are performed.

Note that noise detection for speech recognition is performed by interrupt processing as shown in steps 200 to 202 in FIG.

〔Effect of the invention〕

As explained above, in this invention, the processing after voice recognition is regulated according to the ambient noise.
As the ambient noise increases, the most important processes can be regulated in order, preventing important processes from malfunctioning due to noise.In addition, the display notifies the user of the recognition results, and also controls If the length of the voice phrase is outside the predetermined recognition tolerance range, this will be displayed and the user will be prompted to re-enter the phrase, which will have a significant effect on the ease of use and stable operation of the voice recognition device. .

[Brief explanation of the drawing]

FIG. 1 is a block diagram showing one embodiment of the present invention, FIG. 2 is a block diagram showing essential parts of the apparatus shown in FIG. 1, and FIGS. 3 and 4 are flowcharts showing the operation. 1... Handset, 2... Auto level control device, 3... Voice recognition device, 4... CPU, 5
...Display device, 6...Microphone, 7...Amplifier, 10...Noise level detector, 11a to 11c
... Discriminator, 13 ... Control section.

Claims (1)

[Scope of Claims] 1. A telephone equipped with a voice recognition device that recognizes voice and performs processing according to the recognition result, comprising: a noise level detector that detects an ambient noise level; and a predetermined range for the detected noise level. a discriminator that identifies which category it belongs to; a control unit that regulates the processing content according to the discrimination result; 1. A telephone equipped with a voice recognition device, comprising means for displaying a corresponding display when it is determined that there is a voice recognition device.