JP2006094389A - In-vehicle conversation assisting device - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide an in-vehicle conversation assisting device for enabling occupant of a vehicle to clearly hear conversations at any positions inside an automobile. <P>SOLUTION: A planar type capacitor device array is installed on the ceiling inside the vehicle. A device at a specified position is made to function as a microphone beforehand and a device near a speaker is made to function as a microphone at the time of recognizing a speaking voice. Thus, the speaking voice is surely gathered. Also, output gain adjustment is independently executed to each device made to function as a speaker. Thus, even when the number of the devices made to function as the speaker is increased or decreased, the speaking voice is sounded by a fixed volume at all times. Also, a human detection sensor may be installed inside the vehicle, a device near the occupant inside the vehicle may be made to function as a microphone and a device near a listener may be made to function as a speaker at the time of recognizing the speaking voice. Thus, the conversation is assisted more efficiently. <P>COPYRIGHT: (C)2006,JPO&NCIPI

Description

  The present invention relates to a conversation assistance device in a car, and more particularly to a conversation assistance device that forms a sound field with a microphone and a speaker attached in the vehicle.

  In recent years, passenger cars are equipped with devices that generate sound such as car audio and car navigation, and there is a situation in which the conversation between passengers cannot be heard. Also, while driving on highways, the engine noise of the car, road noise from the tires, and wind noise increased, making it increasingly difficult to hear the conversation.

  Therefore, a plurality of conversation assistance devices have been proposed so that the conversation between passengers can be heard clearly.

  Patent Document 1 discloses an in-car conversation assistance device in which a microphone and a seating sensor are installed at each seat in an automobile, and an output level for each speaker is set in accordance with the position of the speaker to change the direction in which the conversational sound is heard. Has been proposed.

Patent Document 2 proposes an automatic gain adjustment device that performs gain control after suppressing noise and can be heard at an appropriate volume according to the distance between the speaker and the microphone.
JP 2002-51392 A JP-A-9-311696

  However, the in-vehicle conversation assisting device described in Patent Document 1 includes a microphone for each seat and detects whether or not an occupant is seated to control the sound field. For example, a flat in which three people can sit in parallel In the case of a seat, since the seating position is ambiguous, appropriate sound field control cannot be performed.

  The automatic gain adjustment device described in Patent Document 2 always supplies a constant sound pressure level to the sound generation part (speaker or the like), and has a problem that it is difficult to hear when the listening side is away from the speaker installation position.

  SUMMARY OF THE INVENTION An object of the present invention is to provide a conversation assisting device that allows a conversation to be clearly heard regardless of the position of an occupant.

  According to the first aspect of the present invention, there is provided a sound collecting element array in which a plurality of sound collecting elements that function as a microphone or a speaker by switching are arranged on the ceiling of a vehicle, a sound collecting element that functions as a microphone, and a speaker. Switching means for independently switching the sound collecting elements to be operated for each sound collecting element, voice recognition means for recognizing conversational sound from the sound collected by the sound collecting elements functioning as microphones, and each seat in the vehicle Instruct the switching means to place a plurality of sound collecting elements near the position or on the ceiling of the vehicle as sound collecting elements that function as microphones, and talk from the sound collected by the sound collecting elements that function as the microphones. When the voice is recognized, it is released to function as a microphone according to the volume of the conversation voice or the recognition position. And control means for instructing the switching means to change the number and arrangement of the sound elements, characterized by comprising a.

  In the present invention, a planar capacitor type device that functions as a microphone or a speaker is installed in an array on the ceiling of an automobile. A plurality of planar capacitor type devices in the vicinity of each seat or in the vehicle are set in advance to function as a microphone. When an occupant in the vehicle has a conversation, the number and arrangement of planar capacitor type devices that operate as a microphone are changed according to the detected volume and position of the speaker.

  The invention described in claim 2 includes a plurality of human sensors for detecting the position of the occupant, and the control means releases a part of the sound collecting elements near the occupant detected by the human sensor as a microphone. The switching means is instructed to change the arrangement to be a sound collecting element, and the switching means is changed to change the arrangement of the sound collecting elements other than the vicinity of the occupant detected by the human sensor to the sound collecting elements that function as speakers It is characterized by instructing.

  In this invention, in the above invention, the quantity and arrangement of the planar capacitor type device that functions as a microphone are changed according to the arrangement of the occupant detected by the human sensor. Therefore, it is possible to reliably collect conversational voice. Further, by operating a planar capacitor type device other than the vicinity of the passenger detected by the human sensor as a speaker, it is possible to efficiently emit sound to the listener.

  The invention according to claim 3 further includes gain control means for controlling the gain independently for each sound collecting element, and the control means makes the sound collecting sound that makes the sound collecting element near the listener function as a speaker. The gain control means is set so as to increase the gain of the sound collecting element in the direction of the speaker.

  In this invention, the gain of each planar capacitor type device is controlled independently in the above invention. The planar capacitor type device near the listener functions as a speaker and increases the gain in the direction of the speaker. As a result, the voice can be heard from the direction of the speaker.

  As described above, according to the present invention, a plurality of planar capacitor devices can be arranged in a panel on the automobile ceiling, and the connection of each device can be switched and used as a microphone or a speaker. And the number and arrangement of speakers can be easily changed according to the position of the speaker or listener, and the conversation voice can be efficiently transmitted to the listener regardless of the position of the speaker. .

Hereinafter, the in-vehicle conversation assistance device of the embodiment of the present invention will be described in detail with reference to the drawings.
FIG. 1 is a schematic diagram of an automobile equipped with the in-vehicle conversation assisting device of the present embodiment. FIG. 1A is a schematic view of the automobile as viewed from the side, and FIG. 1B is a schematic view of the inside of the automobile. As shown in the figure, the planar capacitor device array 1 is installed on the ceiling inside the automobile. The in-vehicle conversation assisting apparatus according to the present embodiment collects a conversational voice of a speaker by using a planar capacitor device array 1 installed on the ceiling, and transmits the conversational voice to a listener by adjusting the gain in the car to generate a sound. Is.

  The interior of the vehicle includes a right front seat S11, a right rear seat S12, a left front seat S21, and a left rear seat S22. An occupant (driver) H11 is seated on the seat S11, and an occupant H12 is seated on the seat S12. Further, the seat is not limited to two rows and two people as shown in the figure, and may be an automobile having more seats.

  FIG. 2 is a diagram showing the configuration of the planar capacitor device array. As shown in the figure, the planar capacitor device array 1 includes n devices D11 to D1n, n + 1 devices D21 to D2n + 1, and n devices D31 to D3n, which are arranged without being spaced apart from each other. And n + 1 devices D41 to D4n + 1. Each device of the planar capacitor device array 1 is rearward in order from the front of the automobile, devices D11, D12,..., D1n in the rows of devices D11 to D1n, and devices D21, D22, D in the rows of devices D21 to D2n + 1. ..., D2n + 1, devices D31 to D3n are arranged in the order of devices D31, D32, ..., D3n, and devices D41 to D4n + 1 are arranged in the order of devices D41, D42, ..., D4n + 1. ing.

  Further, each of the device array row including devices D11 to D1n, the device array row including devices D31 to D3n, the device array row including devices D21 to D2n + 1, and the device array row including devices D41 to D4n + 1 constitute each row. The center position of the device to be shifted is ½ of the length of the device in the longitudinal direction of the automobile. That is, each device is arranged in a staggered pattern in the planar capacitor device array 1 as a whole.

  Here, all the devices except the devices D21, D2n + 1, D41, and D4n + 1 are formed by a substantially square planar capacitor device, and this is a basic shape. On the other hand, the devices D21, D2n + 1, D41, and D4n + 1 are formed in a substantially rectangular shape that is half of the substantially square shape, that is, a half size of the basic shape. The end portions of the devices D11, D21, D31, and D41 in the front direction of the vehicle coincide with each other, and the end portions of the devices D1n, D2n + 1, D3n, and D4n + 1 coincide with the rear portion of the vehicle. Is a rectangle. The rectangular planar capacitor device array 1 is installed on the ceiling of the automobile.

  Further, the planar capacitor device is a sound collecting element that functions as a microphone or a speaker by switching. That is, in the planar capacitor device according to the present embodiment, when a sound signal is input, the diaphragm vibrates due to the suction / repulsion force generated between the electrodes in the internal electrode to generate sound waves. Further, when the diaphragm vibrates due to a sound wave from the outside, an audio signal can be extracted by a change in the capacitance generated by the vibration.

  FIG. 4A is a diagram when the devices D11 to D1n and D41 to D4n + 1 function as microphones, and FIG. 4B illustrates the case where the devices D12, D1n-1, D43, and D4n function as microphones. FIG. One row or a plurality of rows can function as microphones as shown in FIG. 6A, and a device at a specific position can also function as a microphone as shown in FIG. In FIG. 5B, a device near each seat functions as a microphone. Thus, each device can function as a microphone or a speaker by individual switching.

  FIG. 3 is a block diagram showing the configuration of the in-vehicle conversation assisting device of the present embodiment. As shown in the figure, the in-car conversation assistance device includes a planar capacitor device array 1 having the above-described configuration, and switches SW11 to SW4n + 1 respectively connected to the devices D11 to D4n + 1 of the planar capacitor device array 1. A voice recognition unit 3 and a gain adjustment unit 5 that are connected to each switch, and a switching control unit 4 that controls the operation of each switch. The in-car conversation assistance device includes a control unit 2 that controls operations of the voice recognition unit 3, the gain adjustment unit 5, and the switching control unit 4.

  The switches SW11 to SW1n, SW21 to SW2n + 1, SW31 to SW3n, SW41 to SW4n + 1 are each one corresponding to the devices D11 to D1n, D21 to D2n + 1, D31 to D3n, D41 to D4n + 1 constituting the planar capacitor device array 1. Connected. Further, each switch connects each device to either the voice recognition unit 3 or the gain adjustment unit 5 in accordance with a switching instruction from the switching control unit 4. This allows each device to function as a microphone or speaker. The device functioning as a microphone transmits sound waves received from the outside to the voice recognition unit 3 as a voice signal. The device functioning as a speaker outputs the audio signal output from the gain adjusting unit 5 to the outside as a sound wave.

  The voice recognition unit 3 analyzes a voice signal input via a switch from a device functioning as a microphone, and determines whether the input voice signal is a conversational voice. In the analysis method, for example, an input audio signal is compared with an audio signal pattern stored in advance in a memory (not shown) or the like. Here, the comparison method uses a probabilistic model such as a hidden Markov model. By recognizing only the conversation voice from the device functioning as a microphone, it is possible to recognize that a passenger in the car speaks.

  The control unit 2 controls operations of the voice recognition unit 3, the gain adjustment unit 5, and the switching control unit 4 as described above. The control unit 2 causes a device at a specific position to function as a microphone as an initial state of the in-car conversation assistance device. For example, as shown in FIG. 2A, the switching control unit 4 is instructed to cause the devices D11 to D1n and D41 to D4n + 1 to function as microphones. That is, by using a device in which two rows of ceilings in the vehicle function as microphones, it is possible to specify in which region in the vehicle the conversational voice is generated. Further, the control unit 2 instructs the switching control unit 4 to temporarily function the periphery of the device having the highest voice input level as a microphone when the conversation voice is recognized. As a result, the voice of the speaker can be reliably and efficiently collected.

  For example, as shown in FIG. 2B, the switching control unit 4 may be instructed to cause the devices D12, D1n-1, D43, and D4n to function as a microphone. That is, a device near each seat may function as a microphone in advance. Further, the control unit 2 may instruct the switching control unit 4 to temporarily function the periphery of the device that has recognized the conversation voice as a microphone.

  Note that when collecting the conversational voice of the speaker, the sound signal collected by each device is changed by individually changing the sound collection timing of each device so that the directionality of the sound collection is in the direction of the speaker. You may make it synthesize | combine. Thereby, it is possible to reliably collect only the voice of the speaker.

  The gain adjusting unit 5 adjusts the output gain of the sound that is generated by the device functioning as a speaker. The gain adjusting unit 5 can perform gain control independently for each device. The control unit 2 sets the gain adjustment unit 5 to dynamically perform gain adjustment according to the number of devices that function as speakers. For example, when the number of devices functioning as speakers is small, the output gain of each device is increased so as not to decrease the overall output gain. As a result, the conversation voice can be pronounced at a constant volume as a whole.

  Further, the output gain may be adjusted according to the position of the speaker. For example, the output gain is sequentially increased from the device functioning as a speaker near the listener to the direction of the speaker. As a result, the voice can be heard from the direction of the speaker.

  In addition, you may connect so that the audio | voice signal of car audio etc. (not shown) may be input into the gain adjustment part 5. FIG. In this case, when the control unit 2 does not recognize the conversation, the control unit 2 instructs the gain adjustment unit 5 to generate the car audio musical sound from the device functioning as a speaker. When the conversation voice is recognized, the control unit 2 It may be instructed to lower the volume and preferentially pronounce the conversation voice.

FIG. 4 is a block diagram of the in-vehicle conversation assisting device further provided with a human sensor.
This in-car conversation assistance device further includes a human sensor 6 in addition to the in-car conversation assistance device described above. The human sensor 6 detects the presence of an occupant in the automobile. For example, a plurality of infrared sensors are installed in the vehicle to identify the position of the occupant in the vehicle. The human sensor 6 transmits the detected position information of the occupant to the control unit 2. The control unit 2 instructs the switching control unit 4 to change the arrangement of the device that functions as a microphone according to the position information of the occupant. For example, when the human sensor 6 detects the presence of a passenger near the devices D12, D1n-1, D43, and D4n in the planar capacitor device array 1 shown in FIG. 2B, the devices D12, D1n-1, D43 are detected. , And D4n function as microphones. If the occupant moves at this time, the control unit 2 instructs the switching control unit 4 to change the arrangement of devices that function as a microphone. For example, when it is detected that an occupant near the device D43 has moved to the vicinity of the device D33, the device D43 functions as a speaker, and the device D33 functions as a microphone. As a result, it is possible to reliably recognize conversational speech.

  Further, when the conversation voice is recognized in each device, the control unit 2 instructs the switching control unit 4 to cause the device near the speaker to function as a microphone. As a result, the voice of the speaker can be reliably and efficiently collected. In addition, the switching control unit 4 is instructed to cause a device at a position where an occupant other than the speaker is present to function as a speaker. For example, when an occupant near the device D43 recognizes that a conversational voice has been generated, the devices D32, D33, D42, D44 and the like are caused to function as microphones, and the devices D12, D1n, and D4n are temporarily functioned as speakers. As a result, the conversational voice of the speaker can be reliably transmitted to the listener.

FIG. 5 is a flowchart showing the operation of the in-vehicle conversation assisting device in the present embodiment.
First, it is determined whether each device functions as a microphone or a speaker (s1). Here, as described above, all of one row or a plurality of rows may function as a microphone, or a device at a specific position may function as a microphone. Further, in the case of an in-vehicle conversation assisting device provided with the human sensor 6, the arrangement of a device that functions as a microphone is determined according to position information.

  Thereafter, it is determined whether or not the conversation voice is recognized (s2). Here, as described above, an audio signal input from a device functioning as a microphone is analyzed to determine whether it is a conversational voice. When it is determined that the voice is a conversational voice, the arrangement of the device functioning as a microphone and the device functioning as a speaker is changed (s3). In changing the arrangement, a device functioning as a microphone is arranged in the vicinity of the device in which the conversational voice is recognized, and other devices are arranged as devices functioning as speakers. After that, the output gain of the sound to be generated by the device functioning as a speaker is adjusted (s4). At this time, gain control is performed according to the number of devices functioning as speakers. As described above, the output gain of each device is adjusted so that the overall output gain is constant. Further, the output gain may be increased in order from the listener to the speaker. After adjusting the output gain, the conversation voice is pronounced (s5).

  After the pronunciation, it is determined whether or not the conversation of the same speaker is finished (s6), and the gain adjustment and the pronunciation are repeated until the conversation is finished. When it is determined that the conversation is over, the process is repeated from the recognition of the voice.

  As described above, the in-vehicle conversation assisting apparatus according to the present embodiment has a planar capacitor device that functions as a microphone in the vicinity of the seat in advance, and functions as a microphone near the speaker when speech is recognized. By arranging the devices, it is possible to reliably collect conversational voices. In addition, it is possible to reliably transmit the conversation voice to the listener by arranging a planar capacitor device that functions as a speaker except near the speaker. In this way, the device that functions as a microphone and the device that functions as a speaker are dynamically repositioned according to the positions of the speaker and listener so that the conversation is clear regardless of the position of the occupant. It is possible to hear.

Schematic diagram of a car equipped with an in-car conversation assist device Block diagram showing the configuration of a planar capacitor device array Block diagram of in-car conversation assistance device Block diagram of in-car conversation assistance device with human sensor Flow chart showing operation of in-car conversation assist device

Explanation of symbols

1-planar capacitor device 2-control unit 3-voice recognition unit 4-switch control unit 5-gain adjustment unit 6-human sensor H-occupant S-seat SW-switch D-device

Claims (3)

A sound collecting element array in which a plurality of sound collecting elements that function as a microphone or a speaker by switching are arranged on the ceiling of the interior of a car,
A switching means for switching the sound collecting element to function as a microphone and the sound collecting element to function as a speaker independently for each sound collecting element,
A voice recognition means for recognizing a conversation voice from a voice collected by a sound collection element that functions as a microphone;
A plurality of sound collecting elements near each seat position in the car or on the ceiling in the car are instructed to the switching means in advance to be arranged as sound collecting elements that function as microphones, and sound is collected by the sound collecting elements that function as the microphones. Control means for instructing the switching means to change the quantity and arrangement of the sound collecting elements to function as a microphone according to the volume or recognition position of the conversation voice when the conversation voice is recognized from the voice;
An in-car conversation assist device. Equipped with multiple human sensors that detect the position of the passenger,
The control means instructs the switching means to change a part of the sound collecting element near the occupant detected by the human sensor to an arrangement that makes the sound collecting element function as a microphone,
The in-vehicle conversation assisting device according to claim 1, wherein the switching means is instructed to change the arrangement of the sound collecting elements other than the vicinity of the passenger detected by the human sensor to the sound collecting elements that function as speakers. Further comprising gain control means for controlling the gain independently for each sound collecting element;
The control means is a sound collection element that causes a sound collection element in the vicinity of the listener to function as a speaker, and is set in the gain control means so as to increase the gain of the sound collection element in the direction of the speaker. The in-vehicle conversation assistance device according to claim 2.

Images