JP2019159099A - Music reproduction system - Google Patents

Abstract

To reduce user's stress when a music which is not selected is played, when the music is selected only by a voice.SOLUTION: The music reproduction system includes a sound collection unit which collects sound, a voice recognition unit which recognizes voice collected by the sound collection unit, a candidate song extraction unit which extracts a plurality of candidate songs based on music selection information included in the information voice-recognized by the voice recognition unit, a candidate song part reproduction unit which respectively sequentially reproduces a part of the plurality of candidate songs extracted by the candidate song extraction unit, and a selected song reproduction unit which reproduces a song selected from the reproduced candidate songs based on a song selection instruction input included in the information collected by the sound collection unit and recognized by the voice recognition unit for a reproduction of the candidate song part reproduction unit.SELECTED DRAWING: Figure 1

Description

  The present invention relates to a music playback system.

  In recent years, music distribution services via networks have become widespread. In such a music distribution service, a user connects an information terminal (smart phone, PC, audio device, etc.) having a function of reproducing music to the network, and the music information sent from the music distribution server via the network Is downloaded to the information terminal and then played back, or simultaneously downloaded (streamed) while being downloaded to the information terminal (see, for example, Patent Document 1 below).

  Recently, an information terminal having a voice interactive interface has been used as an information terminal on the user side, and a system for searching information in an interactive format has become widespread. Furthermore, fully interactive information terminals called voice interactive speakers (smart speakers, AI speakers, etc.) that have only voice input / output and no manual input functions or display screens are becoming widespread. This voice interactive speaker is a voice recognition information retrieval system that incorporates AI (artificial intelligence), etc., and provides information requested by the user interactively, eliminating cumbersome operation inputs and hands-free. It has the convenience that it can be used in

JP 2004-191701 A

  When the above-described speaker for voice conversation is used as the user side information terminal of the music distribution service, the information of the music that the user wants to listen to is input by voice and sent from the music distribution server to the user side information terminal. The music information is not confirmed by the user until it is reproduced by the speaker for voice dialogue. In the case of an information terminal equipped with a conventional display screen, the user can check the transmitted music information on the display screen. However, a voice interactive speaker that does not have a display screen performs such confirmation. Can not do it.

  Then, there is no problem as long as music information that matches the user's music selection is sent, but when music information different from the user's music selection is sent, the user plays the music different from the music selection. If you continue to listen to it or you stop playing the song and re-input the voice again, and you cannot reach the desired song even after repeated re-inputs, you may not be able to substitute other input means. The user will feel great stress.

  In the above explanation, a case where a speaker for voice conversation is connected to an online music distribution service has been described as an example. However, even in an offline music playback system, it is assumed that hands-free use with only voice input is assumed. The same problem will occur.

  The present invention has been proposed to address such problems. That is, the subject of the present invention is to reduce the user's stress when a music piece that has not been selected is played when the music piece is selected only by voice.

In order to solve such a problem, the present invention has the following configuration.
A plurality of candidates based on a sound collection unit that collects sound, a voice recognition unit that recognizes voice collected by the sound collection unit, and music selection information included in the information recognized by the voice recognition unit A candidate song extraction unit that extracts songs, a candidate song partial playback unit that sequentially plays a part of the plurality of candidate songs extracted by the candidate song extraction unit, and playback of the candidate song part playback unit A selected music reproduction unit that reproduces a music selected from the reproduced candidate music, based on a music selection instruction input included in the information collected by the sound collection unit and recognized by the voice recognition unit; A music playback system comprising:

It is explanatory drawing which showed the music reproduction system which concerns on embodiment of this invention. It is explanatory drawing shown in the music reproduction system which concerns on other embodiment of this invention. It is explanatory drawing which showed the example of the flow of a process with respect to the audio | voice recognized information in a music reproduction system. It is explanatory drawing which showed the example of the method of selecting a selection music from a candidate music ((a) shows the 1st example and (b) shows the 2nd example). It is explanatory drawing which showed the structural example of the music reproduction system.

  Hereinafter, embodiments of the present invention will be described with reference to the drawings. In FIG. 1, the music playback system 1 includes a sound collection unit 10, a voice recognition unit 11, a candidate music extraction unit 12, a candidate music part playback unit 13, and a selected music playback unit 14. This music reproduction system 1 is a system for selecting and reproducing music only by voice input, and it is an object to reduce the stress received by the user at the time of music selection by the user's voice input.

  In the music reproduction system 1, first, the voice recognition unit 11 recognizes the voice collected by the sound collection unit 10. Next, based on the music selection information included in the information recognized by the voice recognition unit 11, the candidate song extraction unit 12 extracts a plurality of candidate songs. Then, the candidate song partial reproduction unit 13 sequentially reproduces the plurality of candidate songs extracted by the candidate song extraction unit 12. At this time, the candidate music part reproduction unit 13 reproduces a part of the music. That is, the candidate music part reproduction unit 13 partially extracts a chorus part, an intro part, etc. from each music piece and reproduces it. Thereafter, when the music selection instruction input is included in the information collected by the sound collection unit 10 and voice-recognized by the voice recognition unit 11 for reproduction by the candidate music partial reproduction unit 13, the candidate is selected by this music selection instruction input. The selected music playback unit 14 plays back the music selected from the music.

  According to such a music playback system 1, when a user selects music by inputting sound into the sound collection unit 10, a plurality of candidate songs are extracted based on music selection information included in the voice-recognized information. The intro part and the chorus part of the candidate music are played sequentially. When there is a song that matches the song selection among the candidate songs, the user inputs a song selection instruction by voice input. Thereby, the music selected by the music selection instruction is reproduced. According to this, compared with the case where a user's music selection is extracted independently, the probability that the music matched with the user's music selection will be included in the candidate music increases, and it becomes possible to reduce the user's stress.

  FIG. 2 shows another embodiment of the present invention. The music playback system 1A in this example includes a normal playback mode 15 that extracts and plays back specific music based on music selection information included in the information recognized by the voice recognition unit 11. The voice recognition unit 11 further includes a mismatch reproduction determination unit 16 that determines that the reproduced music does not match the user's music selection based on the information recognized by the voice recognition unit 11, and the mismatch reproduction determination unit 16 does not match. When the determination is made, the candidate song extraction unit 13 described above extracts candidate songs.

  For example, when the music selection operation based on voice recognition in the normal playback mode 15 is repeated a predetermined number of times or more within a predetermined time, the discrepancy playback determination unit 16 plays back the same music played in the normal playback mode 15, When the input is instructed to be interrupted, or when a set word that prompts re-selection is detected more than a predetermined number of times during a predetermined time, etc. Is activated when it is determined that does not match the user's music selection.

  In such a music playback system 1A, when a user selects music by voice input to the sound collection unit 10, first, music selection information included in the information recognized by the voice recognition unit 11 in the normal playback mode 15. The music is extracted based on the above and the music is reproduced. When the user wants to continue playing the played music, the normal playback mode 15 continues.

  On the other hand, for example, when the user gives a stop instruction by voice input, the mismatch playback determination unit 16 determines the mismatch, and in this case, the music selection information included in the voice-recognized information is included. Based on this, a plurality of candidate music pieces are extracted, and the intro part, the chorus part, etc. of the candidate music pieces are reproduced sequentially. When there is a song that matches the song selection among the candidate songs, the user inputs a song selection instruction by voice input. Thereby, the music is selected and reproduced.

  FIG. 3 shows an example of the processing flow of each unit that processes the speech-recognized information in the music playback system 1A. In step S1, the voice recognition unit 11 determines whether or not there is music selection information. Here, the music selection information corresponds to various information such as a song name, artist name, album name, and compilation. If there is no music selection information (step S1: NO), the state of waiting for music selection is continued.

  When the voice recognition unit 11 determines that there is music selection information (step S1: YES), the music corresponding to the music selection is reproduced in the normal reproduction mode 15 (step S2). In addition, along with the reproduction of the music, the inconsistent reproduction determination unit 16 determines inconsistent reproduction (step S3). In step S3, the song selection action is repeated a predetermined number of times or more in a short time, and the reproduction and interruption instructions for the same song are repeated a predetermined number of times or more. It is determined whether or not the situation is repeated a predetermined number of times or more, and it is determined whether or not the user's selection of music matches the playback music in the normal playback mode. When the user's music selection and the reproduction music match (step S3: NO), the normal reproduction mode (step S2) is continued.

  On the other hand, if it is determined in step S3 that the user's song selection and the playback song do not match (step S3: YES), in step S4, for example, “The desired song cannot be found. It is determined whether or not to extract candidate music. In response to this question, if the answer “No” is recognized as a voice (step S4: NO), normal playback continues (step S2). When the answer “yes” is recognized by voice in response to the question “” (step S4), candidate song extraction unit 12 extracts candidate songs (step S5).

  The process of the candidate music extraction unit 12 in step S5 is based on the music selection information included in the information finally recognized by voice, or based on the music selection information included in the information recognized by voice from repeated voice input. A plurality of candidate songs related to it are extracted.

  At this time, when the music selection information is a song name, for example, when there are different songs with the same song name, all or a part of them, similar song names considering the accuracy of voice recognition, and the like are extracted as candidate songs. Further, when the music selection information is the artist name, the music of the same artist different from the music reproduced in the normal reproduction mode, the music of the artist having the same or similar name, and the like are extracted as candidate music. When the music selection information is an album name or the like, other music in the same album, music in the same name album of a different artist, and the like are extracted as candidate music.

  Then, when candidate music is extracted in step S5, reproduction ordering is set for all or part of the extracted music. The order of reproduction is set by, for example, the order based on the probability based on voice recognition, the order based on information specified in advance by the user, or the order based on the past user selection tendency (history). The

  When the order of the music to be reproduced among the extracted candidate music is set, partial reproduction of the candidate music is performed according to the order (step S6). In the partial reproduction of the music, which part (intro part, chorus part, feature part, etc.) of the music is set in advance, and the partial reproduction is performed according to the setting. When a plurality of candidate songs are sequentially played back, they are played back with an interval between them, or played back with a cross-fade, etc., so that the break of each song is easily understood.

  Then, during or after the partial reproduction of the candidate music, when the user inputs a music selection instruction by voice input (step S7: YES), the selected music is specified from the candidate music and the selected music is reproduced. The unit 14 reproduces the music (step S8), and proceeds to the normal reproduction mode (step S2) as it is. The music selection instruction at this time may be any instruction that can be determined as a selection instruction, such as “That!”, “That song!”, And “OK!”.

  If there is no input of the selected music instruction (step S7: NO), it is determined whether or not there is another candidate music in the extracted candidate music (step S9), and there is another candidate music ( In step S9: YES, the candidate music is changed (step S10), and the candidate music is partially reproduced again (step S6). Further, when there is no other candidate music (step S9: NO), a question such as “Do you want to continue” is asked to confirm the continuation (step S11), and when continuing (step S11: YES) The candidate music is re-extracted (step S5). At this time, when the candidate music is re-extracted, the user may be prompted to input a new extraction element by, for example, asking an additional keyword to be a hint for music selection. And when not continuing (step S11: NO), a music selection process is complete | finished.

  FIG. 4 shows an example of a method for selecting a selected song from a plurality of candidate songs. In the example shown in FIG. 4A, the candidate songs are partially reproduced in order (step S20). After all the candidate songs have been reproduced, a question such as “What was the desired song?” Was asked (step S21). ), The selected song is reproduced according to the answer to the answer (step S22). If there is no answer to the question in step S21, it is confirmed whether there is a candidate song (step S23). If there is a candidate song (step S23: YES), the candidate song is again selected. Playback is continued in order (step S20). If there is no candidate song (step S23: NO), after outputting a sound such as "There is no desired song" (step S24), the process proceeds to the normal playback mode (step S2).

  In the example shown in FIG. 4B, the first candidate song is played only for a certain time (step S30), and a question such as "Would you like it?" Is asked (step S31), which is the desired song. If it can be confirmed, the selected song is reproduced (step S32). If the answer to the question in step S31 indicates that the song is not the desired song, it is checked whether there is a candidate song (step S33). If there is a candidate song (step S33: YES) ), The next candidate song is reproduced only for a certain time (step S30). If there is no candidate song (step S33: NO), after outputting a voice message such as "There is no desired song" (step S34), the process proceeds to the normal playback mode (step S2).

  Such music playback systems 1 and 1A can be used for voice input in offline systems such as home audio and in-vehicle audio, but can be effectively used in online music distribution services. it can.

  FIG. 5 shows a system example when the music playback systems 1 and 1A are employed in an online music distribution service. In this example, a music distribution server 21 that can communicate via the network 20 and a speaker device 30 that supports voice conversation, which is a user terminal device that can transmit and receive information to and from the music distribution server 21 via the network 20 are provided. Yes.

  The speaker device 30 includes a microphone 31 that constitutes the sound collecting unit 10 described above, and a speaker unit 32 that converts a sound signal of the reproduced candidate song or selected song into sound. In addition, the speaker device 30 receives an audio signal picked up by the microphone 31 and outputs an audio signal to the speaker unit 32, and distributes the information output from the information processing unit 33 via the network 20. A communication unit 34 is provided that transmits information to the server 21, receives information transmitted from the music distribution server 21 via the network 20, and sends the information to the information processing unit 33.

  In such a system in which the speaker device 30 and the music distribution server 21 are connected to the network 20, the music reproduction system 1, 1 </ b> A described above is configured such that the music distribution server 21 uses some or all of the components except the sound collection unit 10. The information processing unit 33 in the speaker device 30 and the computer of the information processing unit 33 may be appropriately deployed in one or both. At that time, the music distribution server 21 or the information processing unit 33 includes a voice recognition process executed by the voice recognition unit 11, a candidate music extraction process executed by the candidate music extraction unit 12, and a candidate music executed by the candidate music partial reproduction unit 13. A music playback program that causes a computer to execute partial reprocessing, selected music playback processing executed by the selected music playback unit 14, and mismatch playback determination processing executed by the mismatch playback determination unit 16 is installed. The program is recorded in a computer-readable memory or other recording medium.

  When such music reproduction systems 1 and 1A are used, when music selection is performed using only audio signals, it is possible to reduce the stress on the user when music that has not been selected is reproduced.

  As described above, the embodiments of the present invention have been described in detail with reference to the drawings. However, the specific configuration is not limited to these embodiments, and the design can be changed without departing from the scope of the present invention. Is included in the present invention. In addition, the above-described embodiments can be combined by utilizing each other's technology as long as there is no particular contradiction or problem in the purpose and configuration.

1, 1A: Music playback system,
10: sound collection unit, 11: voice recognition unit, 12: candidate song extraction unit,
13: Candidate music part playback unit, 14: Selected music playback unit, 15: Normal playback mode,
16: Inconsistent reproduction determination unit,
20: Network, 21: Music distribution server,
30: Speaker device (user terminal device), 31: Microphone (sound collecting unit),
32: Speaker unit, 33: Information processing unit, 34: Communication unit

Claims (7)

A sound collection unit for collecting sound;
A voice recognition unit that recognizes voice collected by the sound collection unit;
A candidate song extraction unit that extracts a plurality of candidate songs based on music selection information included in the information recognized by the voice recognition unit;
A candidate song partial reproduction unit that sequentially reproduces a part of each of the plurality of candidate songs extracted by the candidate song extraction unit;
With respect to the reproduction of the candidate song partial reproduction unit, the candidate song is selected from the reproduced candidate songs based on the song selection instruction input included in the information collected by the sound collection unit and recognized by the voice recognition unit. A selected music playback unit that plays the selected music,
A music playback system comprising: Based on the music selection information included in the information recognized by the voice recognition unit, a normal playback mode for extracting and playing back a specific song,
Based on the information recognized by the voice recognition unit, a mismatch reproduction determination unit that determines that the reproduced music does not match the user's music selection,
The music reproduction system according to claim 1, wherein when the inconsistent reproduction determination unit determines that there is a discrepancy, the candidate music extraction unit performs candidate music extraction. A music playback method for selecting and playing music based on a user's voice selection input,
A speech recognition process for recognizing input speech,
A candidate song extraction step for extracting a plurality of candidate songs based on the music selection information included in the voice-recognized information;
A candidate song partial reproduction step of sequentially reproducing a part of each of the plurality of candidate songs extracted in the candidate song extraction step;
A selected song that reproduces a song selected from the reproduced candidate songs based on a song selection instruction input included in information obtained by voice recognition of the user's voice for reproduction in the candidate song partial reproduction step A music reproduction method comprising: a reproduction step. A voice recognition process for recognizing the collected voice;
A candidate song extraction process for extracting a plurality of candidate songs based on music selection information included in the information recognized by the voice recognition process;
A candidate song partial reproduction process for sequentially reproducing a part of each of the plurality of candidate songs extracted by the candidate song extraction process;
For the reproduction of the candidate song partial reproduction process, a song selected from the reproduced candidate songs based on a song selection instruction input included in the information collected and voice-recognized in the voice recognition process A music reproduction program for causing a computer to execute selected music reproduction processing to be reproduced.   A computer-readable recording medium on which the music reproduction program according to claim 4 is recorded. A music distribution server capable of communicating via a network;
A user terminal device capable of transmitting and receiving information to and from the music distribution server via a network,
The user terminal device includes the sound collection unit and a speaker unit that converts the reproduced candidate music or the music to sound,
The music playback system according to claim 1, wherein each unit of the music playback system is provided in one or both of the computer of the user terminal device and the computer of the music distribution server. A speaker device used in the music playback system according to claim 1,
A speaker apparatus comprising: the sound collection unit; and a speaker unit that converts the reproduced candidate music or the music to sound.

Images