TWI416354B - System and method for automatically searching and playing songs - Google Patents

Abstract

A system for automatically searching and playing songs is provided. The system includes: a tune capturing unit for capturing a section of tune sung by a user; a characteristic calculating unit for calculating a characteristic of the tune according to a levinson-durbin recursion algorithm; a relation degree calculating unit for calculating the relation degree between the characteristic of the tune and each characteristic of each of the songs stored in a memory, and searching for a song having a maximum relation degree with the tune; and a playing unit for playing the searched song. A method for automatically searching and playing songs is also disclosed. By using the invention, the speed of searching and playing the songs can be improved greatly.

Description

System and method for automatically searching and playing music

The present invention relates to a system and method for automatically searching and playing music.

Today's mobile phones, in addition to use for calls, other additional features are increasingly rich, mobile phone developers and service providers spare no effort to tap, the entertainment value of mobile phones is increasingly prominent.

With the increasing enrichment of people's lives, mobile phones with simple call functions can't satisfy the diversified and personalized needs of consumers, and entertainment-oriented mobile phones have emerged. And entertainment-type phones are very attractive to younger generations who are increasingly numerous and have potential for consumption. For example, users can use their mobile phones to listen to music.

However, when listening to music using a conventional mobile phone, the user has to manually search for the location of the favorite music and manually turn on the music. This is very inconvenient. Especially when the user is driving, if you manually search for and start playing music, it may be dangerous to the user.

In view of the above, the preferred embodiment of the present invention provides an automatic search and play The music system and method solve the shortcomings of the traditional mobile phone that requires the user to manually search for music and start playing music.

A system for automatically searching and playing music, the system running in an electronic device, the electronic device comprising a memory and a player, the memory for storing music and characteristic parameters of each piece of music, the automatic search and The system for playing music includes: a melody capture unit for capturing a melody of the user singing; a feature parameter calculation unit for calculating a feature parameter of the melody using the levinson-durbin recursion algorithm; and an association calculation unit for After obtaining the characteristic parameters of all the music in the memory, the nearest neighbor rule is used to calculate the correlation between the feature parameters of the melody and the characteristic parameters of all the music in the memory, and search for the feature parameter with the highest relevance to the melody. a music; and a launching unit for initiating the player to play the searched music.

A method for automatically searching and playing music, the method is applied to an electronic device including a memory and a player, the method comprising the steps of: capturing a melody of a user singing after the electronic device is turned on; using a levinson-durbin recursion The algorithm calculates the characteristic parameters of the melody. After obtaining the characteristic parameters of all the music in the memory, the nearest neighbor rule is used to calculate the correlation between the characteristic parameters of the melody and the characteristic parameters of all the music in the memory, and search for and The music corresponding to the feature parameter with the highest melody relevance; and starting the player to play the searched music.

Compared with the prior art, the system for automatically searching and playing music can calculate the characteristic parameters of the melody by capturing the melody of the user singing, and searching for the highest correlation with the above characteristic parameters from the electronic device. Music, and turn on the player to play the music, so as to automatically search and open the music. Therefore, with the present invention, the user can quickly and conveniently select and play the favorite music.

1‧‧‧Mobile phone

10‧‧‧Automatically search and play music systems

12‧‧‧ memory

14‧‧‧ Player

100‧‧‧ melody capture unit

110‧‧‧Characteristic parameter calculation unit

120‧‧‧ Relevance computing unit

130‧‧‧Starting unit

1 is an application environment diagram of a system for automatically searching and playing music according to the present invention.

2 is a functional unit diagram of a system for automatically searching and playing music according to the present invention.

3 is a flow chart showing the operation of the method for automatically searching and playing music according to the present invention.

Figure 4 is a schematic diagram of the association between the characteristic parameters of the present invention.

Referring to FIG. 1, an application environment diagram of a system for automatically searching and playing music according to the present invention is shown. The system 10 for automatically searching and playing music (hereinafter referred to as "the system 10") runs in an electronic device such as a mobile phone, an MP3, or an MP4 that can play music. In the preferred embodiment, the electronic device is a cell phone. The mobile phone 1 also includes a memory 12 and a player 14. The system 10 can calculate the feature parameters captured by the melody using the levinson-durbin recursion algorithm after capturing the melody of the user singing, and calculate the above characteristic parameters and the feature parameters pre-stored in the memory 12 by using the nearest neighbor rule. Correlation, search for the music corresponding to the most relevant feature parameter and play the music. The memory 12 can be a Flash for storing music that the user likes, and in addition to storing characteristic parameters of each piece of music. Among them, the characteristics of each piece of music The number can also be calculated by the levinson-durbin recursion algorithm.

Referring to FIG. 2, it is a functional unit diagram of a system for automatically searching and playing music according to the present invention. The system 10 includes a melody capture unit 100, a feature parameter calculation unit 110, an association calculation unit 120, and a startup unit 130.

When the user wants to listen to music, but does not want to open the music player 14 by manual operation or inconvenient manual operation, the user can sing a melody after starting the system 10. The melody capturing unit 100 captures the melody sung by the user (hereinafter referred to as "the melody of the segment").

The feature parameter calculation unit 110 is configured to calculate a feature parameter of the segment melody by a levinson-durbin recursion algorithm.

The levinson-durbin recursion algorithm calculates the feature parameters by segmenting the melody. The formula is as follows: among them, Represents the calculated characteristic parameter of the melody of the segment; x ( n ) represents the input signal, that is, each segment of the melody obtained by dividing the segment of the melody that the user sings into p segments; p represents the segment in which the melody is segmented The number can be set according to the user's needs, such as 10 segments; the value of n is any value from 2 to p +1, a (2),..., a ( p +1) is the correlation of each small melody The value of the parameter, e ( n ) can be set according to the requirement, for example, it can be set to 0.1. When the value of e ( n ) is determined, a set of the most approximate values a (2),..., a ( p +1) can be finally obtained by the above two formulas.

The association calculation unit 120 is configured to calculate a correlation between a feature parameter of the segment of the melody and a feature parameter of all music in the memory 12 by using a nearest neighbor rule, and determine the relationship with the feature parameter of the segment of the melody. The feature parameter with the highest relevance of the segmental melody.

The formula for calculating the correlation by the nearest neighbor rule is as follows: dist ( x , x _train )=| x - x _train | ² , where x _train represents the characteristic parameter of the music stored in the memory 12, if the mobile phone stores 10 The music will have 10 sets of x _train values; x represents the characteristic parameters of the melody of the user singing. The smaller the value of dist ( x . x _train ), that is, the closer the x _train value is to the x distance, the higher the correlation between the x _train value and the x value. As shown in FIG. 4, the characteristic parameter 41 is a characteristic parameter indicating the melody of the segment, and the others are characteristic parameters for storing music in the memory 12, and the closer the distance to the feature parameter 41 is, the stronger the correlation is.

After the activation unit 130 finds the feature value x _train value with the highest relevance, the startup unit 130 searches for music corresponding to the group x _train value, and starts the player 14 to play the music.

3 is a flow chart showing the operation of the method for automatically searching and playing music according to the present invention. Step S301, after the user turns on the mobile phone, the mobile phone 1 starts to run the system 10. When the user wants to listen to music and sings a melody, the melody capturing unit 100 captures the melody that the user sings.

Step S302, calculating a characteristic parameter of the melody by the levinson-durbin recursion algorithm.

The levinson-durbin recursion algorithm calculates the feature parameters by segmenting the melody. The formula is as follows: among them, Represents the calculated characteristic parameter of the melody of the segment; x ( n ) represents the input signal, that is, each segment of the melody obtained by dividing the segment of the melody that the user sings into p segments; p represents the segment in which the melody is segmented The number can be set according to the user's needs, such as 10 segments; the value of n is any value from 2 to p +1, a (2),..., a ( p +1) is the correlation of each small melody The value of the parameter, e ( n ) can be set according to the requirement, for example, it can be set to 0.1. When the value of e ( n ) is determined, a set of the most approximate values a (2),..., a ( p +1) can be finally obtained by the above two formulas.

Step S303, reading characteristic parameters of all music stored in the memory, all the sounds The characteristic parameters of music can also be calculated by the above-mentioned levinson-durbin recursion algorithm.

Step S304, when calculating the feature parameter of the melody, calculating the correlation between the feature parameter of the melody and the feature parameters of all the music in the memory 12 by using the nearest neighbor rule, and finding the highest correlation with the melody of the segment. Characteristic parameters.

The formula for calculating the correlation by the nearest neighbor rule is as follows: dist ( x , x _train )=| x - x _train | ² , where x _train represents the characteristic parameter of the music stored in the memory 12, if the mobile phone stores 10 The music will have 10 sets of x _train values; x represents the characteristic parameters of the user singing the melody. The smaller the value of dist ( x . x _train ), that is, the closer the x _train value is to the x distance, the higher the correlation between the x _train value and the x value. As shown in FIG. 4, the characteristic parameter 41 is a characteristic parameter indicating the user's humming melody, and the other is a characteristic parameter for storing music in the memory 12, and the closer the distance to the characteristic parameter 41 is, the stronger the correlation is.

Step S305, after determining the feature parameter with the highest relevance to the segment melody, searching for the music corresponding to the feature parameter with the highest relevance, and turning on the player 14 to play the music.

It should be noted that the above embodiments are merely illustrative of the technical solutions of the present invention, and are not intended to limit the present invention. A person skilled in the art should understand that the technical solutions of the present invention may be modified or equivalently substituted without departing from the spirit and scope of the technical solutions of the present invention.

10‧‧‧Automatically search and play music systems

100‧‧‧ melody capture unit

110‧‧‧Characteristic parameter calculation unit

120‧‧‧ Relevance computing unit

130‧‧‧Starting unit

Claims (4)

A system for automatically searching and playing music, the system running in an electronic device, the electronic device comprising a memory and a player, the memory for storing music and characteristic parameters of each piece of music, the system comprising: a melody capture unit for capturing a melody of the user's humming; a feature parameter calculation unit for calculating a feature parameter of the melody using the levinson-durbin recursion algorithm; and an association calculation unit for acquiring all the music in the memory After the characteristic parameters, the nearest neighbor rule is used to calculate the correlation between the characteristic parameters of the melody and the characteristic parameters of all the music in the memory, and search for the music corresponding to the feature parameter with the highest relevance to the melody; and the starting unit, Used to start the player to play the searched music; the characteristic parameters of the melody of the segment are calculated by the levinson-durbin recursion algorithm as follows: among them, Represents the calculated characteristic parameter of the melody of the segment; x ( n ) represents the input signal, that is, each segment of the melody obtained by the user's melody is divided into p segments; p represents the number of segments in which the melody is segmented , can be set according to user needs; the value of n is any value from 2 to p +1, a (2),..., a ( p +1) is the relevant parameter corresponding to each small melody, e ( n The value can be set according to the requirements; the correlation is calculated by using the nearest neighbor method as follows: dist ( x , x _train )=| x - x _train | ² , where x _train represents the music stored in the memory The characteristic parameter; x represents the characteristic parameter that the user sings the melody of the segment. According to the system for automatically searching and playing music according to claim 1, the electronic device is any one of a mobile phone, an MP3 or an MP4. A method for automatically searching and playing music, the method is applied to an electronic device including a memory and a player, the method comprising the steps of: capturing a melody of a user singing after the electronic device is turned on; using a levinson-durbin recursion The algorithm calculates the characteristic parameters of the melody. After obtaining the characteristic parameters of all the music in the memory, the nearest neighbor rule is used to calculate the correlation between the characteristic parameters of the melody and the characteristic parameters of all the music in the memory, and search for and The music corresponding to the feature parameter with the highest melody relevance; and starting the player to play the searched music; the feature parameters of the melody are calculated by the levinson-durbin recursion algorithm as follows: among them, Represents the calculated characteristic parameter of the melody of the segment; x ( n ) represents the input signal, that is, each segment of the melody obtained by the user's melody is divided into p segments; p represents the number of segments in which the melody is segmented , can be set according to user needs; the value of n is any value from 2 to p +1, a (2),..., a ( p +1) is the relevant parameter corresponding to each small melody, e ( n The value can be set according to the requirements; the correlation is calculated using the nearest neighbor method as follows: dist ( x , x _train )=| x - x _train | ² , where x _train represents the characteristic parameters of the music stored in the memory ; x represents the characteristic parameter that the user sings the melody of the segment. The electronic device is any one of a mobile phone, an MP3 or an MP4, according to the method for automatically searching for and playing music according to claim 3 of the patent application.