JPH022157B2 - - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to a voice recognition device for a vehicle in which voice registration is not possible when the noise inside the vehicle is large.

2. Description of the Related Art Recently, voice recognition devices have begun to be used in various industrial fields, which recognize commands and information input by voice and determine whether or not the voice is voice. In this type of speech recognition device, speech is registered in advance and compared with input speech to determine whether it is similar or not.

FIG. 1 is a block diagram of a conventional speech recognition device of this type. In the figure, 1 is a microphone for inputting voice commands, 2 is a voice recognition unit that registers or recognizes voice based on the voice command signal from microphone 1, 3 is a registration switch that puts the voice recognition unit 2 into registration mode, and 4 is a battery. Power supply voltage is supplied from the power supply to the voice recognition unit 2 via the ignition switch.

In the voice recognition device, when the user gets in the vehicle and turns the ignition switch to the ACC (accessory) position, power supply voltage is supplied from the power supply section 4 to the voice recognition section 2 via the diode 5. Thereafter, in order to register the voice, the registration switch 3 is turned on to put the voice recognition unit 2 into the registration mode, and the voice to be registered is input to the microphone 1, so that the voice recognition unit 2 registers the voice.

In such a voice recognition device, in addition to the engine sound constantly being heard while the vehicle is running, sounds from the fan, radio, and audio equipment may also be heard. In addition, when the sunroof or windows are opened while driving, noise from outside the vehicle enters the cabin, so when registering audio, those noises are superimposed on the audio, and the appropriate There is a problem that registration cannot be performed. Such problems occur not only when the vehicle is moving but also when the vehicle is stopped. Also, when registering multiple words, the order of registration is determined in advance, and the words are displayed in order, but in such cases, register while looking at the display panel while driving. Therefore, it is not very good for driving.

The present invention has been made in view of the above points,
In order to prevent voice registration when the noise inside the vehicle is large, the registration operation is prohibited when it is detected that the noise inside the vehicle has increased.

The present invention will be explained below based on the drawings.

FIG. 2 is a block diagram showing an embodiment of the vehicle voice recognition device according to the present invention. In the figure, the same reference numerals as in FIG. 1 indicate the same components, and this embodiment differs from the conventional example shown in FIG. 1 in that an inverter 6 and an AND circuit 7 are provided. One input terminal of the AND circuit 7 is connected to the registration switch 3, the other input terminal is connected to the ignition switch via the inverter 6, and the output terminal of the AND circuit 7 is connected to the voice recognition section 2.

The voice recognition unit 2 includes a registration means for registering the voice input from the microphone 1, and a recognition means for determining whether the voice input from the microphone 1 is the voice registered in the registration means,
The registration means operates in a registration mode, and the recognition means operates in a recognition mode.

Next, the operation of this embodiment will be explained.

First, when the ignition switch is set to the ACC position, power supply voltage is supplied to the voice recognition section 2. At this time, no signal is output from the ignition-on position (hereinafter abbreviated as IGN-ON position), so
The output from the IGN-ON position becomes L (LOW) level and is inputted to one input terminal of AND circuit 7 via inverter 6 as H (High) level signal. Thereafter, when the registration switch 3 is turned on, an H level signal is input to the other input terminal of the AND circuit 7, so that the AND condition of the AND circuit 7 is satisfied and its output becomes H level. As a result, the voice recognition unit 2 enters the registration mode, and by inputting voice to the microphone 1, the voice recognition unit 2 registers the voice.

Next, turn the ignition switch to the IGN-ON position.The output from the IGN-ON position becomes H level and is input to one input terminal of the AND circuit 7 via the inverter 6, so in this state, the registration switch 3 is turned ON. Even if the AND condition of the AND circuit 7 is not satisfied, the speech recognition section 2 is not in the registration mode, and as a result, no speech is produced in the speech recognition section 2.

Turn the ignition switch IGN-ON like this.
In other words, when the engine is running (there is engine noise), voice registration is no longer possible, and the recognition rate can be improved.

FIG. 3 shows another embodiment of the vehicle voice recognition device according to the present invention. In this embodiment, an RS flip-flop 8 and an OR circuit 9 are provided, and the set terminal S of the RS flip-flop 8 is connected to the registration switch 3, and the reset terminal S of the RS flip-flop 8 is connected to the registration switch 3. The terminal R is connected via an OR circuit 9 to a terminal that outputs the IGN-ON position of the power supply unit 4 and a signal indicating that the registration of the voice recognition unit 2 has been completed.
Further, an output terminal Q of the RS flip-flop 8 is connected to the speech recognition section 2.

In this example as well, the ignition switch is
When set to the ACC position, power supply voltage is supplied to the voice recognition section 2. If you turn on registration switch 3 here,
Since the RS flip-flop 8 is set, its Q output becomes H level, and as a result, the voice recognition section 2
enters the registration mode, and registers the voice by inputting the voice to the microphone 1. Thereafter, when the registration is completed in the voice recognition unit 2, a registration completion signal is output, and the RS is sent via the OR circuit 9.
Since flip-flop 8 is reset, its Q
The output becomes L level, and as a result, the voice recognition section 2 ends the registration mode. If the ignition switch is set to the IGN-ON position during registration, OR circuit 9
Since the RS flip-flop 8 is reset via the RS flip-flop 8, the registration mode of the voice recognition section 2 is interrupted and voice registration ends midway, but the voice is not registered. Next, when the ignition switch is set to the IGN-ON position before voice registration, the RS flip-flop 8 is reset via the OR circuit 9, so even if the registration switch 3 is turned on afterwards, the voice recognition section 2 will not be registered. mode and cannot register audio.

If the ignition switch is set to the IGN-ON position during voice registration in this way, the registration mode ends at that point and no further registration is possible. Furthermore, since the registration mode ends each time voice registration is completed, even if the noise level inside the vehicle becomes high after the window is opened, it will not be possible to register the sound.This improves the recognition rate even more than the embodiment shown in Fig. 2. be able to.

In the above-mentioned embodiment, due to the engine noise when the engine is activated by the ignition switch,
We have explained the situations in which the noise level inside the vehicle increases, but the noise level inside the vehicle also increases when the fan, radio, or audio system is running, or when the sunroof or window is opened. Therefore, the present invention may detect that the noise level inside the vehicle has increased and prohibit the registration operation. In addition, if the item that increases the noise level mentioned above is activated by supply from the battery by detecting the state of the ignition switch, it is sufficient to detect whether the ignition switch is ON or OFF, but a sunroof that opens and closes manually In the case of windows and windows, it may be detected from their movement that the noise level inside the vehicle has increased.

In the vehicle voice recognition device mentioned above, there are restrictions on where the microphone can be installed, so it is usually mounted on a sun visor,
The roof, spot lights, etc. must be installed at a considerable distance (30 to 80 cm) from the passenger's mouth, which can generate a lot of noise when the vehicle is driving through tunnels or at high speeds. If voice recognition is performed in such a case, the noise may cause erroneous recognition and reduce the recognition rate. Therefore, it is preferable to provide an additional microphone in addition to the microphone provided at the above position. FIG. 4 is a block diagram of an embodiment of a speech recognition device provided with such two microphones. In the figure, reference numeral 10 denotes a microphone mounted in a position far away from the passenger's mouth, such as on the sun visor, roof, or spot light.
1 is an amplifier that amplifies the audio signal 100 from the microphone 10; 12 is an auxiliary microphone that is attached to a position near the passenger's mouth, such as at the collar of a passenger's jacket; 13 is an amplifier that amplifies the audio signal 120 from the auxiliary microphone 12. amplifier, 14 is amplifier 1
Signal 110 from 1 and signal 13 from amplifier 13
A changeover switch 15 for switching between the changeover switch 14 and the changeover switch 14 is a voice recognition unit that recognizes voice based on the signal 140 from the changeover switch 14. 5 shows an embodiment of the changeover switch 14 shown in FIG.
6 and auxiliary microphone plug 17 and auxiliary microphone plug 1
7, and a contact 18 which is turned ON and OFF by the switch 7.

Next, in the embodiment shown in FIG. 4, when performing voice recognition during normal driving, the auxiliary microphone plug 17 is pulled out from the jack 16 as shown in FIG. 5A. At this time, since the contacts 18 are in contact as shown, the terminals S and S' are electrically connected. When audio is input into the microphone 10 in this way, the audio signal 100 from the microphone 10 is amplified by the amplifier 11 and becomes a signal 110.
The signal is input to the voice recognition section 15 as a signal 140 by the changeover switch 14, so that voice recognition is performed.

Next, when performing voice recognition when the noise is relatively high, such as when driving in a tunnel or at high speed, insert the auxiliary micro plug 17 into the jack 16 as shown in Figure 5 (b), and use the auxiliary microphone. 12 is fixed to the collar of the passenger's jacket with a clip or the like. At this time, the contacts 18 are separated as shown, so that the terminals S and S' are electrically cut off. Here, when audio is input to the auxiliary microphone 12, the audio signal 120 from the auxiliary microphone 12 is
is amplified by the amplifier 13 to become a signal 130, which is input to the speech recognition section 15 as a signal 140 by the changeover switch 14, so that speech recognition is performed.

In this way, since the auxiliary microphone 12 is located closer to the passenger's mouth than the microphone 10, the gain of the amplifier 13 is set smaller than the gain of the amplifier 11. Therefore, when voice is input using the auxiliary microphone 12 when the noise is high, the voice recognition unit 15
The recognition rate improves because the noise level input to the image becomes smaller.

As explained above, in the present invention, the registration operation is prohibited when it is detected that the noise inside the vehicle has increased, so that appropriate registration can be performed without being affected by noise. It is possible to improve the recognition rate. According to the present invention, since there is no need to perform registration while driving, there is no need to look at the display panel for registration while driving, which is favorable for driving.

[Brief explanation of drawings]

Figure 1 is a block diagram of a conventional speech recognition device.
FIG. 2 is a block diagram showing one embodiment of the vehicle voice recognition device according to the present invention, FIG. 3 is a block diagram showing another embodiment of the vehicle voice recognition device according to the present invention, and FIG. A block diagram showing an embodiment of a vehicle voice recognition device equipped with two microphones,
FIG. 5 shows an embodiment of the changeover switch shown in FIG. 4. 1, 10, 12...microphone, 2, 15...
...Speech recognition section, 3...Registration switch, 4...Power supply section, 6...Inverter, 7...AND circuit, 8...
...RS flip-flop, 9...OR circuit, 11,
13...Amplifier, 12...Auxiliary microphone, 1
4...Switch, 16...Jack, 1
7...Auxiliary microphone plug.

Claims (1)

[Claims] 1. voice input means; display means for displaying words to be registered in a predetermined order when registering voice;
a voice registration command means that is operated when registering a voice and outputs a voice registration signal; a registration means that registers voices in the order in which they are input from the voice input means when the voice registration signal is input; In a vehicle voice recognition device having a recognition means for determining whether an input voice is a voice registered in the registration means, the detection means outputs an on-state signal when an on-state of an ignition switch is detected. and registration inhibiting means for inhibiting the registration operation of the registration means when the on-state signal is input.