JP2006092430A - Music reproduction apparatus - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To automatically reproduce music suitable for user situations and the atmosphere that an environment indicates without troublesome operations. <P>SOLUTION: A keyword detection part 2 detects prespecified keywords from user conversation contents, and an emotion information detection part 3 analyzes prosody information extracted from the user conversation contents to detect user emotion information. A user situation estimation part 4 estimates user situations from the detected keywords and the emotion information, and an atmosphere estimation part 5 classifies a current atmosphere into either of a plurality of categories based on objective information in a car cabin and the user situations, which are collected by an atmosphere factor information collection part 6. A category list, in which various kinds of music data stored in a music database 7 are linked to each classification category to be classified, is stored into a memory 8. A retrieval processing part 9 retrieves music data, which belongs to a category classified by the atmosphere estimation part 5, from the category list, and then retrieves the music data from the music database 7. A speaker unit 10 reproduces the music data. <P>COPYRIGHT: (C)2006,JPO&NCIPI

Description

  The present invention relates to a music playback device that performs automatic playback of music.

Patent Document 1 provides a music data distribution system that enables a user to easily select and play a desired piece of music. In this system, when a user inputs a desired song name or artist name through a microphone on the client terminal side, the voice input signal is transmitted to the music distribution server, and the music distribution server performs voice recognition processing on the received signal. The song name list or artist name list is searched based on the result of performing and the search result list is returned to the client terminal. On the client terminal side, the returned list is displayed, and the music selected by the user's operation from the list is downloaded from the music distribution server and played back.
JP 2002-341880 A

  In recent car navigation systems, a device equipped with a music playback device using a music server is provided, and if such a configuration as in Patent Document 1 is adopted, an added value can be improved. However, even when such a configuration is adopted, since a conscious operation by the user is required for selecting a song, it is impossible to reproduce a desired music piece when the user does not want to perform a conscious operation.

For example, depending on the situation involved in traffic jams or the situation of the driver who is dealing with this situation, it may be frustrating and difficult to listen to music, so music to calm yourself By selecting and playing a song, the consciousness of alleviating this situation also declines, and there is a tendency that the mood becomes increasingly frustrating.
The present invention has been made in view of the above circumstances, and is capable of automatically reproducing a musical composition with a high stage effect suitable for the atmosphere in the environment estimated from the user's situation and the environmental state. The object is to provide a music playback device.

  According to the first aspect of the invention, the voice recognition means detects a preset keyword from the user's utterance content, and the emotion analysis means detects the prosody information (accent, intonation, rhythm, pose, etc.) from the user's utterance content. Information on the prosody of the user), and the user's emotion information is detected by analyzing the prosody information, and the user's situation is set based on the detected keyword and emotion information by the situation estimation means. Classify one of the situation patterns.

  The atmosphere factor information collecting means collects information indicating objectively the environment where the user is placed as atmosphere factor information, and the atmosphere estimation means is configured to determine the atmosphere according to the environment where the user is placed as the situation estimation means. Is classified into one of a plurality of categories set in advance based on the situation pattern classified by the above and the atmosphere factor information collected by the atmosphere information collecting means.

  The storage means stores a category list in which various types of music data stored in the music database are classified in association with the above categories, and the search means is classified according to the atmosphere level from the category list. Music data belonging to the selected category is retrieved, and the music data is acquired from the music database and output. The reproduction means reproduces the music data output in this way.

  Therefore, the music suitable for the user's situation and environment is automatically reproduced without a troublesome operation. For example, as a classification category based on the estimation result of the atmosphere estimation means, it is estimated that the atmosphere is in a situation where the user is frustrated from factors such as weather, noise, or words that the user unconsciously utters, and the corresponding category If the music data corresponding to the category is set to have an effect to alleviate the feeling of irritation, music suitable for relaxing the mood while avoiding the increase in the user's feeling of irritation Will be played.

  According to the invention of claim 2, in the above invention, the atmosphere factor information collecting means is configured to listen in the vehicle interior of an automobile or the like, and the atmosphere is intended for the vehicle interior as an environment where the user is placed. The factor information is configured to collect information such as destination, current location, season, time, weather, vehicle speed, number of brakes, number of signal stops, and the like. For example, if it is estimated that there is an atmosphere in which the user is unconsciously uttering words that indicate unpleasant emotions while being stuck in traffic jams, etc. The music is selected and played to alleviate this. As a result, the feeling of irritation of the user during driving is alleviated and the user feels relaxed, and the dangerous driving caused by the feeling of irritation can be suppressed.

  According to the invention of claim 3, in each of the above inventions, when the user's situation presents a pleasant emotion, a positive emotion, or when the user presents an unpleasant emotion such as frustration or anger When the atmosphere is classified as a negative emotion, the atmosphere estimation means categorizes the atmosphere of the environment where the user is placed into a category, and if that category belongs to a negative emotion, the negative emotion is relaxed and positive. The music data that tends to be directed to the emotions of the music is associated, so at least in the atmosphere where the user is in a negative emotional state, music that can alleviate this can be provided, so that the selection of music is distracted In any case, it can work to recover this automatically. Especially when driving a car, even if the driving user feels frustrated or uncomfortable with the atmosphere in the passenger compartment and tends to drive recklessly, this should be shifted to relieve this. Will be able to.

  According to the invention of claim 4, in each of the above inventions, the music database is a music data server accessible via a communication line, so that the music data server constituting the music database can be shared by a plurality of music playback devices. Therefore, it is useful for realizing cost reduction.

  According to the invention of claim 5, in each of the above inventions, the music data that does not correspond as a category is set, so that it is possible to prevent a situation in which music unnecessary for the user is played.

(First embodiment)
A first embodiment in which the present invention is applied to a music playback device mounted on a vehicle will be described below with reference to FIG.
FIG. 1 shows the overall configuration of a music playback device by a combination of functional blocks. In FIG. 1, a microphone 1 is installed at a place where a user (vehicle driver) can efficiently receive a voice, and an input voice signal is used as a keyword recognition unit 2 and an emotion analysis unit. Is given to the emotion information detection unit 3.

  In this case, the keyword detection unit 2 always performs voice recognition processing based on the voice input signal from the microphone 1, and has a function of detecting a preset keyword from the user's utterance content recognized by the processing. . As keywords at this time, linguistic information that the user is likely to utter unconsciously according to the psychological state (mental situation) or physical condition (physical situation) that changes according to the driving state or environment at that time, For example, languages that can be defined as negative emotions that indicate annoyance or discomfort such as “Osoi”, “Irritation”, “Juutai (congestion)”, “Hayak (early)”, “Madakana”, “Aa”, etc. Linguistic information that can be defined as positive emotions that indicate a good mood state such as “Yoshi”, “Yatta”, “Butterfly”, etc. is set.

  The emotion information detection unit 3 extracts prosody information (information on prosody such as accent, intonation, rhythm, and pose) from the voice input signal from the microphone 1 and analyzes the extracted prosody information to analyze user emotion information. It is the structure which always performs the process which detects this. That is, the emotion information detection unit 3 represents emotion information indicating a user's psychological state from non-verbal information (speech characteristics such as sound intensity, length, height, tone, and pause) included in the user's unconscious utterance. It is the composition which obtains. In this case, as emotion information indicating the user's psychological state, for example, strong verbal words corresponding to the negative emotion described above, non-linguistic information such as sigh, and conversely a song or whistle corresponding to a positive emotion There is non-linguistic information.

  The user situation estimation unit 4 as the situation estimation means estimates the user's mental situation and physical situation based on the keyword detected by the keyword detection unit 3 and the emotion information detected by the emotion information detection unit 3, and the The estimation result is classified into any of a plurality of predetermined categories, and category information indicating the classification result is output to the atmosphere estimation unit 5. The classification category in this case is, for example, a situation corresponding to negative emotions such as “angry”, “uncomfortable”, “irritated”, “sleepy”, “tired”, “bad feeling”, etc. Situations corresponding to positive emotions such as “Fun”, “Relax”, “Good mood” are set.

  The atmosphere estimation unit 5 as an atmosphere estimation unit is a vehicle interior based on the user situation data estimated by the user situation estimation unit 4 and the atmosphere factor information input from the atmosphere factor information collection unit 6 as the atmosphere factor information collection unit. Estimate the atmosphere. In this case, the atmosphere factor information collection unit 6 sets the atmosphere factor information for a plurality of items, and sets the item of the atmosphere factor information in order to estimate the atmosphere in the vehicle interior during driving from various angles. Yes.

FIG. 2 shows the items and types of atmosphere factor information. For example, the items are “present location”, “destination”, “season”, “time”, “weather”, “vehicle speed”, “brake count”, “ Information such as “the number of times of signal stop” is obtained based on information obtained from various sensors provided in the vehicle, a navigation device, or the like.
Here, “current location” or “destination” refers to the geographical environment where the user is currently traveling, or the geographic environment where the destination is going to be. It is intended to analyze the factors that lead to the atmosphere in the vehicle interior. In addition, information such as “season”, “time”, and “weather” is used to analyze the state of the environment outside the vehicle due to the influence of natural changes. The information such as “vehicle speed”, “brake count”, or “signal stop count” is used to analyze, as an element, how the environment related to the driving of the vehicle that the user drives is.

  The music database 7 is composed of a large-capacity storage medium such as a hard disk device, and stores music data corresponding to various types of music. The stored music data includes data stored in advance at the time of purchase, data directly stored from various music media, and data acquired using functions such as communication from the outside. In addition, when storing music data, it is possible to reduce the amount of data by using an audio information compression technique (for example, MP3). Note that this music database may be dedicated, or may be configured using a music data server for a car navigation system mounted on a vehicle.

  The memory 8 as a storage means is constituted by a rewritable nonvolatile memory (or a battery-backed RAM) typified by a flash memory, and the music data stored in the storage device 7 is used as the atmosphere estimation unit 5. A category list is stored that is classified in association with each category to be estimated. In this case, each category is associated with music data for at least one piece of music that is considered to be able to improve the mental situation and physical situation of the user classified into the category.

  Specifically, for example, music data having a user's arousal effect (music having a large change in volume, music having a beat, etc.) is associated with the category “sleepy”, and music data having a user's mood-enhancing effect ( It is conceivable that uptempo music or the like is associated with the category “tired” and music data (for example, healing music) having a user's relaxation effect is associated with the category “irritated”.

  In addition, the classification category by the user situation estimation unit 4 is also set to have no corresponding music data. For example, if a user's mental and / or physical situation is classified into a category corresponding to the positive feeling of “relaxed”, playing music to the user can be counterproductive. For this category, it can be considered that the corresponding music data does not exist.

  The search processing unit 9 that receives the category information output from the atmosphere estimation unit 8 searches the category list in the memory 8 for music data belonging to the category indicated by the category information, and acquires the music data from the storage device 6. And output to the speaker unit 10 as a reproducing means. At this time, if a plurality of music data exists in the category, randomly selected music data is output.

The speaker unit 10 includes a music playback unit 10a including a decoder and an amplifier necessary for playback of normal music data and music data in a compressed state, and a speaker 10b disposed in the passenger compartment. The music corresponding to the given music data is reproduced by the speaker 10b.
Next, the operation of the above configuration will be described. When the vehicle is running on a car, the environment in which it is placed changes variously. And the change of the environment may affect a user's psychological situation etc., and may influence a driving state by extension. This system provides music that automatically determines the atmosphere and relaxes it so that the atmosphere in the vehicle interior does not adversely affect driving.

The operation process is as follows. Now, for example, the operation at that time will be described on the assumption that a running automobile is in a state of being involved in a traffic jam. First, in a traffic jam, a user who is a driver may unconsciously utter “Osoi” or the like, sigh, or have a bitter word.
Since a keyword set in advance is registered, the keyword detection unit 2 detects the keyword when a word uttered by the user is obtained by voice recognition. As keywords, for example, linguistic information (“nemui”, “nemutai”, “nemutsu”, “tsukareta”, “kutabireta”, “arua” that users are likely to utter unconsciously according to changes in their psychological state and physical condition. (Sigh) ”,“ hot ”,“ good ”,“ cursie ”, etc.) are set, and it becomes possible to detect the user's psychological state and physical condition through the keyword detection unit 2.

  Here, for example, it can be determined that the user encounters a traffic jam and is frustrated in response to detection of keywords such as “Osoi” and “Madaka”. In addition, it can be determined that the user is drowsy in response to the detection of keywords such as “nemui”, “nemutai”, “nemutsu”, and the like, “Tsukaleta”, “Kutabileta”, “Aaa (sigh) It can be determined that the user is in a tired state in response to detection of a keyword such as “)”.

On the other hand, since the prosody information is extracted from the utterance content of the user and the emotion information detection unit 3 that detects the emotion information of the user by analyzing the prosody information is provided, the nonverbal included in the unconscious utterance of the user Emotion information indicating the user's psychological state can be obtained from the information (speech characteristics such as sound intensity, length, height, tone, and pause).
The user situation estimation unit 4 estimates the user's mental situation and physical situation based on the keyword detected by the keyword detection unit 2 and the emotion information detected by the emotion information detection unit 3, and the estimation result is preset. It is classified into one of a plurality of categories (“irritated”, “sleepy”, “tired”, “tired and sleepy”, “not feeling good”, etc.).

  Next, the atmosphere estimation unit 5 can obtain the atmosphere factor information for the items shown in FIG. 2 described above from the atmosphere factor information collection unit 6 as objective information according to the driving state of the automobile. Then, the present vehicle interior atmosphere is estimated from the atmosphere factor information and the situation information corresponding to the user situation described above. This is to determine to some extent whether the user situation estimated by the user situation estimation unit 4 is objectively linked to the situation in the vehicle interior. It is intended to determine the atmosphere in the car by having an atmosphere classified as.

  The memory 8 stores a category list in which various types of music data stored in the storage device 6 that is a music database are associated with the above categories, and the search processing unit 5 stores the category list. Music data belonging to the category classified by the atmosphere estimation unit 5 is retrieved, and the music data is acquired from the storage device 6 and output to the speaker unit 8. The speaker unit 8 reproduces the music data output in this way through the speaker 8b.

  Thereby, the music suitable for the user's situation and the atmosphere of the environment is automatically reproduced without a troublesome operation. Here, as a classification category based on the estimation result of the user situation estimation unit 4, a category of “irritated state” indicating a feeling that the user is frustrated by being involved in a traffic jam is set, and the mood is expressed as music data corresponding to the category. Since those having the effect of relaxing are set, music suitable for eliminating the atmosphere of increasing the irritability of the user is reproduced.

  Contrary to the above, if a category of “good mood” is set as a classification category based on the current location, the geographical environment of the destination, etc., and a situation that indicates a good state of the user, it corresponds to that category. As a result, music suitable for the atmosphere of the environment is reproduced. For example, if you are driving along the coast or approaching the coast, a bright song is chosen for the sea, and if you are heading to the ski area, you will be comfortable with snow and skiing. Settings such as selecting a song are made.

  In this case, a category for which the corresponding music data does not exist (for example, a category in which the user's mental situation and / or physical situation is “relaxed”) is also set as the classification category by the user situation estimation unit 4. Therefore, in a situation where it is not necessary for the user to provide a song that has been selected automatically, it is possible to prevent inadvertently providing music and causing adverse effects.

  According to this embodiment, since the atmosphere in the passenger compartment can be estimated by the atmosphere estimation unit 5 and automatically switched to music corresponding to the atmosphere, the user's situation or atmosphere can be in a congested state. It has the effect of suppressing emotions such as frustration. In addition, since such a frustrating mood can be alleviated, it is possible to avoid a situation that tends to be a dangerous driving and to lead to a relaxed mood. Furthermore, since the music suitable for the atmosphere can be switched and provided according to the destination and the weather conditions at that time, the atmosphere in the driving state can be more effectively produced.

  Also, when the user's situation is negative emotion, the atmosphere in the vehicle interior can be relaxed to relax the user's mood, and when the user's situation is positive emotion, It is possible to create a more entertaining environment by further boosting. On the other hand, when the user has a positive emotion, it is possible to respond appropriately by not automatically providing music.

It can be further customized by providing a user with a mode for switching to an appropriate song in various situations. Further, by providing a learning function, music suitable for the user can be provided in a timely manner.
It is also possible to allow the user to program the category and the corresponding song. In this case, it may be possible to cope with a simple operation without being complicated.

  In the present embodiment, the music database 7 is directly connected to the search processing unit 9, so that the search database 5 can directly access the music database 7. For this reason, the speed at which the music data is acquired is increased, and the response until the music reproduction is improved.

(Second Embodiment)
FIG. 2 shows a second embodiment of the present invention. The difference from the first embodiment is that music data is taken in by communication from the outside instead of the music database 7.

  In FIG. 2, a communication unit 11 that can communicate with the outside is connected to the search processing unit 9. The communication unit 11 is configured to be accessible to the music data server 13 via a communication network 12 such as a wireless communication network and the Internet. The music data server 13 is linked with a music database 14 similar to that described above, and provides music data requested in response to an external access.

In the configuration shown in FIG. 2, the music database 7 (see FIG. 1) on the music playback device side shown in the first embodiment is omitted, but the music database 7 may be provided at the same time. .
According to the present embodiment having such a configuration, the music data server 13 including the music database 14 can be shared by a plurality of music playback apparatuses, which is beneficial in realizing cost reduction. Further, for the user, if the capacity on the music data server 13 side is secured, a wide range of latest music can be provided without restriction without providing a large-capacity music database 7 on the car side. It becomes like this.
Note that the music playback method can be played with all the music data received, or so-called streaming playback.

(Other embodiments)
The present invention is not limited to the above embodiment, and can be modified or expanded as follows.

  The user situation estimation unit 4 is configured to estimate both the mental situation and physical information of the user, but may be configured to be able to estimate at least one of the mental situation and physical information. When a positive emotion of “fun” is detected, the music is not played. However, when such a positive emotion is detected, a music that has an effect of sustaining or further enhancing the emotion may be played. . Although an example in which the present invention is applied to a music playback device mounted on a vehicle has been described, the present invention can be applied to music playback devices in general.

In addition to automobiles, the present invention can also be applied to various music playback devices such as home music playback devices and music playback devices installed in environments such as halls where many people gather.
The atmosphere in the passenger compartment becomes a different category depending on the situation of the user who is using it even in the same objective environment, and it becomes tiresome to do the same according to the situation of the user over time, so learning It is also possible to provide a function or to prepare a pattern of various corresponding modes in advance so that it can be selected.

As a learning function, for example, the category adopted in a certain scene and the reaction when the music data is played back are fed back in a good direction or it is not so effective and added to the next judgment function. Is to provide such a function. Thereby, automatic music selection suitable for the user's situation and the atmosphere in the passenger compartment can be provided.
Also, the music associated with the category may be designated by the user according to his / her preference. In this case, it is possible to provide functions such as deleting a default music piece or adding a user-designated list to a default list.

1 is an overall functional block diagram showing a first embodiment of the present invention. Diagram showing the contents of multiple categories to be classified FIG. 1 equivalent diagram showing a second embodiment of the present invention

Explanation of symbols

1 is a microphone, 2 is a keyword detection unit (speech recognition unit), 3 is an emotion information detection unit (emotion analysis unit), 4 is a user situation estimation unit (situation estimation unit), 5 is an atmosphere estimation unit (atmosphere estimation unit), 6 is an atmosphere factor information collection unit (atmosphere factor information collection unit), 7 and 14 are music databases, 8 memories, 9 is a search processing unit (search unit), 10 is a speaker unit (reproduction unit), 11 is a communication unit, 12 Is a communication network, and 13 is a music data server.

Claims (5)

Voice recognition means for detecting a preset keyword from the user's utterance content;
Emotion analysis means for extracting prosody information from the user's utterance content and analyzing the prosody information to detect the user's emotion information;
A situation estimation means for classifying a user situation into any of a plurality of preset situation patterns based on the detection outputs of the voice recognition means and the emotion analysis means;
Atmosphere factor information collection means for collecting information that objectively shows the environment where the user is located as atmosphere factor information;
Any of a plurality of categories set in advance based on the situation pattern classified by the situation estimation unit and the atmosphere factor information collected by the atmosphere information collection unit according to the environment where the user is placed Atmosphere estimation means to classify into,
A music database that stores music data corresponding to many types of music,
Storage means for storing a category list obtained by classifying the music data in association with each category;
Search means for searching for music data belonging to the category classified by the atmosphere estimation means from the category list, and acquiring and outputting the music data from the music database;
Playback means for playing back music data from the search means;
A music playback device comprising: The music playback device according to claim 1,
The atmosphere factor information collecting means is intended for the interior of the vehicle as the environment where the user is placed, and the atmosphere factor information includes destination, current location, season, time, weather, vehicle speed, number of brakes, number of signal stops, etc. Configured to collect information on
The music reproduction device characterized in that the atmosphere estimation means classifies the atmosphere in the vehicle compartment into any of the plurality of categories. The music playback device according to claim 1 or 2,
The atmosphere estimation means is a positive emotion when the user's situation is presenting a pleasant emotion, and a negative emotion when presenting an unpleasant emotion such as being frustrated or angry, When the atmosphere of the environment in which the user is placed is classified into a category, if the category belongs to the negative emotion, the negative emotion is relaxed and the music data having a tendency toward the positive emotion is associated. A music playback device characterized by that. The music playback device according to any one of claims 1 to 3,
The music player is a music data server accessible via a communication line. The music playback device according to any one of claims 1 to 4,
A music playback apparatus characterized in that a category for which no corresponding music data exists is set in the category.