JP2007043356A - Device and method for automatic sound volume control - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a device capable of controlling output of on-vehicle information equipment so that conversation can be held normally, since it is difficult to decide whether the conversation is made normally with in-cabin environmental parameters for making a conversation effectively in a cabin although a unit is disclosed which controls sound volume of the on-vehicle equipment, in response to the speech volume of a speaker or specified sound of the siren of an ambulance etc., as the in-cabin environmental parameters. <P>SOLUTION: The device is constituted so that words for posing a question back at or confirming words of the opposite side have been registered previously in conversation contents of users; and when these registered words are contained in the conversation content, it is decided that the conversation is not being conducted effectively, so that the sound volume of the on-vehicle information equipment is controlled as a background sound to the previously-stored level. <P>COPYRIGHT: (C)2007,JPO&INPIT

Description

  The present invention relates to an automatic volume control device for an in-vehicle device such as a car radio or a car stereo that is a source of ambient noise so that a conversation between passengers in a moving body such as a vehicle is not hindered.

  In recent vehicles, comfort has been improved and the noise level in the vehicle has been well controlled. At the same time, sounds generated by in-vehicle acoustic devices such as car radios and car stereos have become more problematic than engine sounds, tire running sounds and other outside vehicle noises during running. In the above-described in-vehicle device, a passenger who is a user can freely adjust the volume, but on the other hand, the volume is constant unless the user adjusts the volume. For this reason, when a conversation starts in the vehicle, the volume of these in-vehicle acoustic devices that did not get in the way before that may interfere with the conversation.

Thus, it is necessary to pay attention to a specific information source (sound source) or to control the volume of these in-vehicle acoustic devices depending on the background noise. For example, in the following (Patent Document 1), a voice information signal and other acoustic signals serving as background sounds are separated and received from the acoustic information in the vehicle, and a conversation between passengers in the vehicle is performed. In other words, a method is disclosed in which the volume of the background sound is adjusted by determining that the conversation is in progress if the voice information signal is maintained at a predetermined level or higher. In addition, a device that lowers the volume of in-vehicle audio equipment when detecting the sound of emergency vehicles such as sirens and railroad crossing alarms (Patent Document 2), or increases the volume of in-vehicle information equipment when the noise level in the vehicle increases. In addition, a volume control device (Patent Document 3) for an in-vehicle acoustic device that does not respond to voice information, does not increase the volume, and does not hinder passengers' conversation is also disclosed.
JP 2004-48570 A Japanese Patent Laid-Open No. 11-136059 Japanese Patent Laid-Open No. 7-50533

  As described above, the automatic volume control of the in-vehicle acoustic device that reacts only to the alarm sound for emergency or the in-vehicle noise level is not effective as a measure for reducing the interference with the conversation in the vehicle. This is because the average voice level varies from person to person, and since there is a small voice, there is no conversation, and because the speaker is speaking at a loud volume, the volume of the in-vehicle acoustic device is not necessarily high. If used, there may be inconveniences. Therefore, in the present invention, the conversation between the passengers is not determined only by the volume, but the volume of the conversation is determined not by the volume level but by the parameter obtained from the conversation contents, and the volume automatic adjustment that correctly corresponds to the in-vehicle situation. The purpose was to realize the device.

  In the present invention, the sound component in the vehicle is recognized and processed for each of the two components of the sound signal component due to the in-car conversation and the noise component from the in-vehicle information providing device as the background sound, and the sound signal component is heard back, or If the speaker's words are included, it is judged that the conversation between the users who are passengers has not been established, and the information providing the noise source so that the conversation between the users can be established An automatic volume control apparatus and method capable of adjusting the volume from an apparatus.

  In the present invention, it is possible to determine whether or not a conversation is established from the utterance content, not just the volume, and to control the volume of the background sound that becomes noise according to the user's condition. Talking in an environment that matches

(Embodiment 1)
Hereinafter, the present invention will be described with reference to the drawings. FIG. 1 is a block diagram showing a basic configuration of a volume automatic adjusting apparatus according to the present invention. This volume automatic adjusting device is composed of two parts, a sound signal analysis adjusting unit 01 and an information providing device 02. The sound signal analyzing and adjusting unit 01 collects a vehicle interior sound, and outputs a sound signal and other noise signals. The signal is separated, the respective signals are analyzed, and the volume control signal is generated. The information providing apparatus 02 corresponds to an in-vehicle acoustic device, and is a part that adjusts the volume by the volume adjustment signal and outputs an acoustic signal.

  In FIG. 1, a sound collecting unit 10 collects in-vehicle sounds including voice and noise together, and this output signal is simultaneously input to the voice recognition unit 11 and the noise recognition unit 16. The voice recognition unit 11 recognizes the conversation content in the input voice signal and determines the volume of the speaker's voice. Data for determining whether or not to adjust the volume of the in-vehicle acoustic device is recorded in the volume adjustment state storage unit 13 in advance. The volume adjustment determination unit 12 generates a volume adjustment signal from the recognition / determination result while referring to the signal from the volume adjustment state storage unit 13.

  On the other hand, the noise recognizing unit 16 extracts a noise component excluding the voice signal from the inputted in-vehicle sound, and the noise level detecting unit 17 measures the noise level. The measured noise level data is input to the volume adjustment storage unit 14 and the volume adjustment learning unit 15. Here, the volume adjustment storage unit 14 is a portion that stores the value of the amount of adjustment of how much the volume is reduced with respect to the input noise level, how much the volume is reduced by the information providing device 02 that is an in-vehicle acoustic device. The stored value is input to the volume control unit 20 of the information providing device 02, and is also input to the volume control learning unit 15 of the sound signal analysis control unit 01 at the same time.

  The volume adjustment signal generated by the volume adjustment determination unit 12 and the volume adjustment value that is output from the volume adjustment storage unit 14 are input to the volume adjustment unit 20 of the information providing device 02, and the output volume is adjusted to make the speaker unit. 22 is output. Here, the volume value immediately before the volume adjustment is performed is stored in the volume storage unit 21. This stored value is stored in the volume adjustment value learning unit 15 of the volume automatic adjustment device 01 for use as reference data in the next volume generation signal generation.

  The signal processing procedure of the volume automatic adjusting apparatus having the configuration shown in FIG. 1 will be described with reference to the flowchart of FIG. First, the sound collecting unit 10 collectively collects in-vehicle sounds including voice and noise (step S100). The collected signal is input to the speech recognition unit 11 and the noise recognition unit 16, and speech recognition is first performed as signal processing. In order to perform voice recognition, it is determined whether or not a voice signal is included in the collected signal. Here, if the audio signal is not included, it is not necessary to adjust the volume, and the process ends (step S110 / NO).

  When the audio signal component is included (step S110 / YES), the volume adjustment determination unit 12 determines whether or not the conversation is established. For this determination, it is necessary to adjust the volume from the information providing apparatus 02 by referring to a word registered in the volume adjustment state storage unit 13 in advance, or by determining whether the same word is being heard back. Judgment is made. Therefore, for example, words as shown in Table 1 are stored in the volume adjustment state storage unit 13.

この登録されている言葉に該当する言葉が音声信号中に含まれておれば、会話として成立していないことを示しているから、情報提供機器の音量を下げる必要があることを示す音量調節信号を音量調節判定部１２で発生させる（ステップＳ１２０／ＹＥＳ）。また、このような言葉が含まれていない場合は、搭乗者間での会話が成立していることを示しており情報提供機器の音量を調節する必要がないから処理は終了する（ステップＳ１２０／ＮＯ）。

If a word corresponding to this registered word is included in the audio signal, it indicates that the conversation is not established, so a volume adjustment signal indicating that the volume of the information providing device needs to be lowered Is generated by the volume adjustment determination unit 12 (step S120 / YES). Further, when such a word is not included, it indicates that a conversation between the passengers has been established, and the process ends because there is no need to adjust the volume of the information providing device (step S120 / NO).

In the above processing, when it is determined that the volume adjustment of the information providing device needs to be executed (step S120 / YES), first, the noise recognition unit 16 recognizes the noise component of the in-vehicle sound, and then the noise amount N _t. Is measured by the noise detection unit 17 (step S130). When the volume adjustment signal generated by the volume adjustment determination unit 12 indicates that the volume adjustment is necessary, the volume adjustment value V _t suitable for the noise amount N _t stored in the volume adjustment value storage unit 14 in advance. Until the sound volume adjustment unit 20 of the information providing apparatus 02 reduces the sound volume (step S140). In the above description, voice recognition is performed first and then noise detection is performed. However, this order may be reversed. Further, a plurality of levels of the volume adjustment value V _t may be prepared for the representative noise amount N _t .

Now that lower the volume, speaker learns below the optimum volume control value V _t to the operator or the like of the聞者or the apparatus (step S150~S160). This learning ends with the end of the conversation. However, the end of the conversation is recognized as the end of the conversation when the presence of the conversation is not recognized for a predetermined time (step S160 / YES) and stored in the volume storage unit 21. The volume of the information providing apparatus 02 is returned to the volume V _current before the volume adjustment, and the automatic volume adjustment ends.

FIG. 3 shows the contents of the processing procedure in the learning. The learning of the present invention is to cause the automatic volume control device to learn the optimal volume control value V _t for a speaker or a listener. That is, when the conversation is started between the passengers, either the volume of the information providing apparatus 02 is increased or decreased so that the speaker, the listener, or the operator of the apparatus can smoothly advance the conversation. Is performed, it is determined that the output volume of the information providing apparatus 02 is not appropriate, and the volume adjustment value V _t is adjusted up or down in accordance with the direction of the operation so as to be an optimum value. . The value of the volume control value V _t If the above operation volume control is not performed the value of the volume control value V _t as was the appropriate hold it. Note that the volume control value V _t can be stored by setting a plurality of volume control values V _t as unique volume control values V _t for a plurality of users.

  The learning process in the present invention will be described below with reference to FIG. FIG. 3 shows the contents of step S150 “learning process” in FIG. In FIG. 3, when the conversation between the speakers or the passengers who are the listeners is first started, the speaker, the listener or another third party operates the volume control unit 20 of the information providing device 02 to change the volume level. Is performed (step S200 / YES), it is determined in which direction the level change is made to increase or decrease the volume (step S210). If the volume adjustment by the user is not performed (step 200 / NO), the learning process (step S150) is not necessary, and the data passes as it is.

Here, when it is adjusted towards lowering the volume (step S210 / YES) is judged to be a level at which it interferes with a conversation level of current volume adjustment value V _t is too high, the volume control value V _t The value is set to a newly operated smaller value (step S230). Conversely, when it is adjusted towards increasing the volume (step S210 / NO), for example in such cases the topic for the output of the information providing apparatus 02 determines that the value of the volume control value V _t is too small operation 3 is set to a value obtained by increasing this value to the sound volume level, the learning process in FIG. 3 is terminated, and the flow chart of FIG.

Allowed to learn the value of the set volume control value V _t as described above as an output volume level of the information providing apparatus 02 during subsequent vehicle conversation. This long as the subsequent conversation continues value of the volume control value V _t is maintained (step S160 / NO), for example, the output is the conversation if no output predetermined time or more ends from the voice recognition unit 11 (Step S160 / YES), the output volume of the information providing device 02 is returned to the same volume V _current state as that in the no-talking state before the adjustment by the volume adjustment unit 20, and a series of processing relating to the volume adjustment is performed. finished, thereafter, when the conversation in the car has occurred using the value of the learned volume control value V _t.

Further, a plurality of volume adjustment values V _{t1 to} V _tm are set for a plurality of users, and identification codes such as user names corresponding to the respective volume adjustment values are assigned to the volume adjustment value storage unit 14. (Step S240), and when performing a conversation, the user can have a conversation in an optimal acoustic environment by first inputting an identification code corresponding to the user.

  FIG. 4 is a block diagram showing a configuration of a volume control device used in the conventional example described in Patent Document 1. In FIG. In FIG. 4, the vehicle interior sound average value is obtained, and a change in the average sound volume is detected by the conversation state detection unit 45 at predetermined time intervals. Here, if the average volume increases, the volume of the conversation increases, and it is determined that the conversation is playing, and the information sound control unit 47 blocks information from the information source 40 that becomes the background sound.

  According to this configuration, since only the increase / decrease in the volume is used as the parameter for the background sound control, individual differences in the volume of the spoken voice are not taken into account, and the conversation is not necessarily played because the volume is high. Further, according to this conventional example, when it is determined that the conversation is playing, the sound of the information source that is the background sound is blocked. For this reason, there is a practical inconvenience, for example, that it is impossible to have a conversation based on the sound of an information source such as news. In the present invention, it is determined whether or not a conversation is established from the content of the user who is a speaker or a listener, and since the configuration is such that the predetermined volume is reduced without blocking the background sound, the practical range is limited. Can be spread.

The block diagram which shows the structure of the volume automatic adjustment apparatus by this invention. The flowchart which shows the process sequence in the volume automatic adjustment apparatus by this invention. The flowchart which shows the learning process of the setting level at the time of the noise amount adjustment used as a background sound. The block diagram of the vehicle-mounted information provision apparatus currently disclosed conventionally.

Explanation of symbols

01: Automatic volume control device 02: Information providing device 10: Sound collection unit 11: Speech recognition unit 12: Volume adjustment determination unit 13: Volume adjustment state storage unit 14: Volume adjustment value storage unit 15: Volume adjustment value learning unit 16: Noise recognition unit 17: Noise level detection unit 20: Volume adjustment unit 21: Volume storage unit 22: Speaker unit 40: Information source 45: Conversation state detection unit 47: Information sound control unit

Claims (6)

A volume automatic control device composed of a volume automatic control device and an information providing device for controlling a sound signal in a passenger compartment,
The volume automatic control device is a volume adjustment state storage means for storing in advance a specific word for detecting that the conversation is not established from the content of the conversation in the vehicle interior;
Volume adjustment value storage means for outputting a volume adjustment value for adjusting the volume when it is detected that a specific word stored in the volume adjustment state storage means is included in the content of the conversation in the vehicle interior And
The information providing device includes a volume adjusting unit that adjusts a volume of the vehicle-mounted information providing device by a predetermined amount according to a volume adjustment value from the volume adjustment value storage unit;
An automatic volume control device characterized by comprising: The volume automatic adjustment device according to claim 1,
When the user manually adjusts the volume after the volume adjustment by the volume adjustment means, the volume adjustment value storage means for storing the volume data after the adjustment is provided, based on the volume data at the next volume adjustment. Automatic volume control device characterized by adjusting the volume. The automatic volume control device according to claim 1 or 2 is an automatic volume control device configured by a volume automatic control device and an information providing device for controlling a sound signal in a vehicle interior,
The volume automatic control device includes:
A sound collecting means for collecting an acoustic signal in the passenger compartment;
Voice recognition means for extracting a voice signal component from the acoustic signal collected by the sound collection means;
Volume adjustment state storage means for storing specific words for detecting that the conversation is not established;
Detecting whether or not a specific word stored in the volume adjustment state storage means is included in the audio signal component, thereby determining whether or not volume adjustment of the information providing apparatus is possible;
Noise recognition means for extracting a noise component as a background sound;
A noise amount detecting means for detecting a noise amount of the extracted noise component;
Volume adjustment learning means for learning an externally input volume adjustment value;
Volume adjustment value storage means for storing a volume adjustment value for adjusting the noise level with respect to the noise level obtained by the noise level detection means, and the information providing apparatus comprises:
A volume control means for manually inputting a volume control signal from the signal from the volume control determination means and the volume control storage means, or by the user manually from the outside, and adjusting the volume according to those signals;
A volume storage means for storing a volume value immediately before the volume is adjusted;
A speaker unit that outputs acoustic information;
The output of the volume storage means is an automatic volume adjustment apparatus connected to the volume adjustment value learning means in the automatic volume control apparatus. The automatic volume control device according to claim 3,
The specific word stored in the volume control state storage means for detecting that the conversation is not established is a plurality of specific words indicating that the listener of the conversation listens back to the speaker. Automatic volume control device. A volume automatic adjustment method for controlling a volume from an information providing device which is a noise component which is a background sound with respect to a conversation voice level in a vehicle interior by a volume automatic control device,
A specific word for detecting that the conversation is not established is previously stored in the volume adjustment state storage means,
The speech recognition means recognizes the conversation component in the passenger compartment, and the volume adjustment determination means determines whether or not the specific word stored in the volume adjustment state storage means is included in the conversation component. ,
If the specific word is not included, the process is terminated without performing the following process.
When the specific word is included, the volume adjustment determination means generates a volume adjustment signal that the volume adjustment is necessary, and inputs the volume adjustment signal to the volume adjustment means included in the information providing device,
On the other hand, the noise component is recognized by a noise recognition means, and the noise amount of the noise component is detected by a noise amount detection means.
The volume adjustment value stored in the volume adjustment value storage means for the noise amount is input to the volume adjustment means of the information providing apparatus together with the volume adjustment signal,
In the volume control means, output from the speaker unit at either the volume set by the volume control signal and the volume control value or the volume set by the volume information manually input by the user from the outside,
At this time, the volume storage unit of the information providing apparatus stores the volume value immediately before volume adjustment, and temporarily stores the volume information manually input by the user from the outside, and inputs it to the volume adjustment value learning unit A volume automatic adjustment method, characterized in that the volume adjustment value can be changed by doing The volume automatic adjustment method according to claim 2,
When storing the externally input volume information in the volume adjustment value learning means, a personal identification code corresponding to the individual user is given and learned and stored in the above,
A method for automatically adjusting a volume, wherein the personal identification code is input during use to enable use at a volume adjustment value suitable for the user.

Images