JPS59180599A - Voice recognition controller to be carried on vehicle - Google Patents

Abstract

(57) [Summary] This bulletin contains application data before electronic filing, so abstract data is not recorded.

Description

DETAILED DESCRIPTION OF THE INVENTION The technical field BA relates to an on-vehicle voice recognition control device, and particularly to prevention of erroneous recognition due to in-vehicle noise.

BACKGROUND ART In recent years, with the rapid development of electronic technology, voice recognition technology has rapidly developed, and vehicle-mounted voice recognition control devices using this voice recognition technology have been proposed. This in-vehicle voice recognition control device (d) provides voice instructions for controlling switches such as car radio and lights, and controlling seats up and down, and is suitable for controlling in-vehicle devices while driving.

FIG. 1 is a block diagram showing an example of an in-vehicle voice recognition control device, in which a voice command issued by a speaker is converted into a 'Kamekisou' by a microphone 1 and supplied to a part of the filter. Ru. 7 irta part 2 b, microphone l
After amplifying the audio signal supplied from the audio signal generator 3, the noise component is removed and only the pure audio signal component is extracted and supplied to the audio recognition section 3. The speech recognition section 3 converts the speech signal supplied from the filter section 2 into a data group having a predetermined number of nodes or less by applying conversion processing such as Fourier transform and Hadamard transform. This data group is then compared with a pre-stored data group, and the most similar data group is selected to recognize the voice signal.

Here, the degree of similarity between the stored data group and the data group obtained by converting the input audio signal is expressed by a constant called the g-distant distance, which is obtained by performing a certain arithmetic operation.

Then, if this recognition distance is less than a predetermined value and is the smallest compared to others, the memory data group and g voice 1 g
The data group whose code has been converted is recognized as the same data group. Data groups recognized as the same in this way are
The contents represented by the stored data groups determined to match are supplied to a control section 4 constituted by a microcomputer. The control section 4 receives the signal supplied from the speech recognition section 3 and controls the corresponding control section 5 . For example, three types of words such as "radio, light, seat" are stored in advance in the speech recognition unit 3 as a data group. Next, when a voice command "radio" is given to the microphone 1, this voice is converted into an electrical signal by the microphone 1, and then supplied to the voice recognition unit 3 via the filter part 2, where it is sent as data. converted to a group. And this data group corresponds to the memorized radio, lights, and seats.
The recognition distance for the data group of the three buildings is determined by comparing it with the data group of eii, and it is determined that the data group is the same as the data group with the smallest recognition distance, and the recognition result is output. However, if all the recognition distances are equal to or greater than a predetermined value, it is determined that there is no corresponding data.

In this manner, when the voice recognition section 3 recognizes that the input voice is "radio", this recognition result is supplied to the control section 4. The control unit 4 is provided with one control signal for each voice input. For example, when a voice input of "radio" is input, the radio switch is turned on.
When you input "Light" by voice, the Hendride switch is turned on. The operation of the speech recognition section 3 is shown in a flowchart as shown in FIG.

However, in the in-vehicle voice recognition control device with the above configuration, it is safe to increase the sensitivity of the microphone in order to distinguish and extract the voice of the speaker from the noise inside the car, but conversely, it is necessary to increase the sensitivity of the microphone. Conversely, there is a problem in that malfunctions may occur due to other people's voices or unrelated audio signals.

DISCLOSURE OF THE INVENTION An object of the present invention is to provide an in-vehicle voice recognition control device that prevents malfunctions due to in-vehicle noise.

In order to achieve such an object, the in-vehicle voice recognition control device according to the present invention uses a first
Speech recognition is performed by selecting the output signal of the microphone, and when the output signal of the first microphone exceeds a predetermined level, the output signal of the second microphone with higher sensitivity is selected and speech recognition is performed. This is what we do.

Therefore, according to the in-vehicle voice recognition control device configured in this way, after detecting the voice uttered by the speaker, the output signal of the second microphone with high sensitivity is selected to perform voice recognition. This has an excellent effect of preventing malfunctions caused by noise inside the car and other people's voices.

BEST MODE FOR CARRYING OUT THE INVENTION FIG. 3 is a circuit diagram showing an embodiment of a vehicle-mounted voice recognition control device according to the present invention. - In the figure, 10.11 is the first. As a second microphone, the first microphone 10 has low sensitivity, and the sensitivity of the second microphone 11 is set easily. 12 is the first. A selector that selects the output signal of the second microphone 10 and 11, and at 2 m always selects and outputs the output of the first microphone 10 (No. W). 13 is a part of the filter, and the output of the selector 12 The noise component included in the signal is removed to extract and output only pure voice, and a control signal is generated when the input signal rises above a predetermined level. 14 is a filter part 13 This is a voice recognition section that performs voice recognition using the input signal as input, and has the same configuration as that shown in FIG.

15 is a system configured by a microcomputer +1
11g15, when the control 1 signal A is supplied, the select signal B is generated and the selector 12 is controlled to select and output the cutting signal of the second microphone 11. In addition, the ``fttl'' part 15 is the voice recognition part 1.
The control target unit (not shown) is controlled according to the g-identification result of step 4.

In the vehicle-mounted voice recognition control device configured as described above, in a normal state, the selector 12 selects and outputs the output signal of the first microphone 10. in this case,
Since the sensitivity of the first microphone 10 is set low, it does not detect the two voices inside the car or the voices of other people under normal conditions, and therefore the voice recognition unit 14 may malfunction due to these noises. This prevents

Next, when the voice-operated speaker, for example the driver, issues a voice command such as "U-Geo", the first. Since the second microphone 10.11 is located close to the driver, the output signal of the first microphone 10 is also increased and supplied to the filter part 13. Filter part 1:3
frJ: Even if the average level of the input signal is detected,
When the predetermined level is exceeded, a control signal A is generated and supplied to the control section 15. When the control section 15 is supplied with the control signal A, the control section 15 generates a select signal B and supplies it to the selector 12, thereby converting the input signal Mlb from the output signal of the first microphone 10 to the output signal of the second microphone 11. Switch to Ig.

As a result, the signal output from the second microphone 11, which has high sensitivity, is supplied to the filter part 13 via the selector 12, and the noise component is removed. Only the voiced words are supplied to the spoken word sewing section 14.

Therefore, in this case, the highly sensitive second microphone 11
In order to detect the voice uttered by the speaker, the difference in level between the speaker's voice No. 48 and the noise and the other person's speaking voice No. 13 becomes large, and unintended signals are removed. This voice signal is then recognized by the voice recognition section 14 in the same manner as in the case of FIG. 1 described above, and the recognition result is supplied to the control section 15. The control unit 15 executes various control operations of the controlled units shown in the figure in accordance with a control answer that is predetermined for each recognition result.

Then, when the voice command by the speaker ends, the filter section 13 stops generating the control signal A because the average input signal level of the filter section 13 falls below a predetermined level.

Therefore, the control unit 15 also controls (stops the generation of the select signal B as the generation of the signal A stops. As a result, the selector 12i"l: stops the capture of the input signal from the second microphone 17.
The output of the first microphone 11 is switched to the output of the first microphone 10, and a standby state is entered.

As explained above, the in-vehicle voice recognition control device with the above configuration uses a 2A (microphone) with different sensitivity for detecting the speaker's voice command and for determining the content of the voice command. Malfunctions caused by other people's voices are prevented, and reliable instructions can be given by voice.

[Brief explanation of the drawing]

Figure 1 is a circuit diagram of an example of an in-vehicle voice recognition control device.
FIG. 2 is a flowchart for explaining the operation of the voice recognition section shown in FIG. 1, and FIG. 3 is a circuit diagram showing an embodiment of the vehicle-mounted voice recognition control device according to the present invention. 10.11...1st and 2nd microphones, 12...
- Selector, 13... Part of filter, 14... i-voice recognition section, 15... Control section.

Claims (1)

[Claims] (1) A first microphone, a second microphone having higher sensitivity than the first microphone, and the first microphone. A selector that selects and outputs the output signal of the second microphone and an output signal of the MiJ selector are input, and noise components are removed to output only the audio signal component, and the average input signal level is predetermined. a part of the filter that emits a control signal when the level exceeds a certain level; a voice recognition unit that recognizes the audio signal output from this part of the filter and generates a recognition signal; and a control signal that is generated from the part of the filter. control to switch the input selection of the selector from the output of the first microphone to the output of the second microphone in response to the received voice recognition unit, and control the corresponding controlled unit in accordance with the recognition signal generated from the voice recognition unit. An in-vehicle voice recognition control device comprising: