JP2004117632A - Play list generator, music player, music delivery system, method therefor, program, and recording medium - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To enable a user to consecutively play the most suitable musical pieces in accordance with his or her image without requiring prior knowledge of musical pieces. <P>SOLUTION: A play list generator is provided with an image transition input part for inputting transition information indicating transition of images of a desired play list, a music source storage part having a plurality of musical pieces stored, an impression value calculation part for extracting musical feature quantities of musical pieces to calculate impression values, a music data storage part having feature quantities and impression values stored, a distribution analysis part for analyzing a transition range indicating a distribution of impression values, an image transition correction part for correcting transition information in accordance with the transition range, a music/vocal discrimination part for discriminating whether musical pieces are music or vocal, a play list generation part for generating a musical piece list, and a play list substitution part for changing the order of musical pieces constituting the musical piece list so as to approximate an impression value list of musical pieces constituting the musical piece list to that of transition information. Thus the user can generate a list of musical pieces to be played, which includes new musical pieces, without requiring prior knowledge when registering them. <P>COPYRIGHT: (C)2004,JPO

Description

[0001]
TECHNICAL FIELD OF THE INVENTION
The present invention relates to a playlist generation device, a music reproduction device, a music distribution system and a method, a program, and a recording medium for continuously providing optimum music according to a user's image.
[0002]
[Prior art]
2. Description of the Related Art Conventionally, as a service for providing continuous music, there are a playlist creation service and a background music providing service. These services ask playlist users individually about their requirements, and experts use songs that match the requirements from a music database based on their musical knowledge and list them in the order of playback. Provided to the public.
[0003]
On the other hand, the applicant generates a playlist by digitizing an image of a certain song as an impression value as an impression value and selecting the song so as to match the transition of the impression value according to the requirement input by the user. A playlist generation device has been disclosed (see Patent Document 1). By adopting this method, the user can obtain a playlist suitable for his desired image without requiring specialized knowledge.
[0004]
[Patent Document 1]
Japanese Patent Application No. 2002-071094
[0005]
[Problems to be solved by the invention]
However, in the following cases, a playlist desired by the user may not be generated.
[0006]
First, the user has generated a playlist on the assumption that the user has a certain amount of music in all images. In other words, there is no problem if a certain number of songs exist near the transition of the impression value according to the requirements, but if the user has a biased music, there is a song that realizes the transition of the impression value. Therefore, there was a possibility that sufficient satisfaction could not be obtained.
[0007]
Next, when audio signals such as English conversation and reading are registered in a music database, these audio signals are also included in the playlist. Since a normal playlist is a list of music, an audio signal is not considered to be appropriate as one music constituting the playlist.
[0008]
When a song to be actually included in the playlist is selected from the transition of the impression value, an impression value (that is, a search center coordinate value) serving as a search center is determined, and a song near that value is selected. However, depending on the distribution of impression values, the order of the song impression values may not be changed in the middle of the playlist, but may be changed in the middle of the playlist. FIG. 2 is a diagram showing a case where the order of the music images is changed. In FIG. 2, the nearest music piece is selected from the search center impression value. At this time, since the distribution of the music is biased, the order of the transition specified by the user and the impression value of the music is reversed when the second music and the third music are selected. In this way, even when the nearest song is selected, the order of the image of the song may be changed in the middle, and when the user uses the rules and history information when selecting the song, the order is further increased. The possibility of replacement is increased. When such a phenomenon occurs, the user looks as if the tune has not been selected according to the transition designated by the user.
[0009]
The present invention solves the above-described problem, and eliminates the need for the user to have prior knowledge of music, and provides a playlist generation device, a music playback device, and a music device that continuously provide optimal music according to the image of the user. It is an object of the present invention to provide a music distribution system, a method thereof, a program, and a recording medium.
[0010]
[Means for Solving the Problems]
The playlist generation device of the present invention is a playlist generation device that generates a playlist indicating the order of songs to be played by the playback device, and a playback condition input unit that inputs a desired playlist length. An image transition input unit for inputting transition information indicating a transition of an image of a desired playlist, a music source storage unit in which a plurality of songs are stored, and a musical feature amount of the song is extracted. Selecting an impression value calculation unit for calculating an impression value indicating an image when listening to music, a music data storage unit storing feature amounts and impression values, and a plurality of songs stored in the music data storage unit A playlist generating unit that generates a song list indicating a list composed of a plurality of songs, and a first impression value list indicating transitions of impression values of the songs that constitute the song list. To approximate the first impression value list showing transition of impression values, characterized in that it comprises a playlist replacement unit to rearrange the order of the songs that constitute the music list.
[0011]
In the playlist generation device according to the present invention, the playlist generation unit generates a third impression value list indicating transition of the impression value using the reproduction condition and the transition information, and the impression value approximate to the third impression value list. Characterized in that the music data storage unit is provided with an impression value search unit for selecting a plurality of tunes giving the transition from the music data storage unit.
[0012]
The playlist creation device according to the present invention includes a music / speech determination unit that determines whether a song is music or voice using an impression value.
[0013]
The playlist generation device of the present invention includes a distribution analysis unit that calculates a transition range indicating a distribution of impression values of songs registered in a music data storage unit, and an image transition correction unit that corrects transition information according to the transition range. It is characterized by having.
[0014]
The music reproducing apparatus according to the present invention is further provided with a music source acquiring section for acquiring music sources of a plurality of songs described in a song list from a music source storage section, and a reproducing apparatus for reproducing a music source.
[0015]
A playlist generation method according to the present invention is a playlist generation method for generating a playlist indicating the order of songs to be reproduced by a reproduction device, wherein: a reproduction condition input step of inputting a desired playlist length; An image transition input step of inputting transition information indicating transition of an image of a playlist to be extracted, and an impression of extracting a musical feature amount of a song and calculating an impression value indicating an image when listening to music from the feature amount. A value calculation step and a playlist generation step of generating a song list indicating a list composed of a plurality of songs by selecting a plurality of songs stored in a music data storage unit in which feature amounts and impression values are stored. And a first impression value list indicating the transition of the impression value of the song constituting the song list, and a first impression value list indicating the transition of the impression value of the transition information. To approximate the to; and a playlist replacement step of rearranging the order of the songs that constitute the music list.
[0016]
In the playlist generation method according to the present invention, in the playlist generation step, a third impression value list indicating transition of the impression value is generated using the reproduction condition and the transition information, and the impression value approximate to the third impression value list is generated. Characterized in that the music data storage unit in which impression values are stored is selected from a plurality of music pieces giving the transition of the impression value.
[0017]
The playlist generation method according to the present invention includes a music / voice determination step of determining whether a song is music or voice using an impression value.
[0018]
The playlist generation method according to the present invention includes a distribution analysis step of calculating a transition range indicating a distribution of impression values of songs registered in a music data storage unit in which impression values are stored, and correcting transition information according to the transition range. And an image transition correcting step.
[0019]
The playlist generation method according to the present invention is characterized by further including a music source acquisition step of acquiring music sources of a plurality of songs described in the list from a music source storage unit, and a reproduction step of reproducing the music sources.
[0020]
A program according to the present invention is a program for causing a computer to execute a playlist generation process for generating a playlist indicating the order of music to be reproduced by a reproduction device, and the generation process includes a desired playlist length. Inputting a reproduction condition, inputting transition information indicating a transition of an image of a desired playlist, and extracting a musical feature amount of a song, and listening to music from the feature amount. An impression value calculating step of calculating an impression value indicating an image of the time; and a list composed of a plurality of songs by selecting a plurality of songs stored in a music data storage unit in which feature amounts and impression values are stored. A playlist generating step of generating a song list indicating the following, and a first impression value list indicating transition of impression values of the songs constituting the song list: To approximate the first impression value list indicating the transition of the impression value of transfer information, characterized in that it comprises a playlist replacement step of rearranging the order of the songs that constitute the music list.
[0021]
The recording medium of the present invention is a computer-readable recording medium that records a program for causing a computer to execute a playlist generation process for generating a playlist indicating the order of songs to be reproduced by the reproduction device, The processing includes a reproduction condition input step of inputting a desired playlist length, an image transition input step of inputting transition information indicating a transition of a desired playlist image, and a musical feature amount of a song. Extracting and calculating an impression value indicating an image when listening to music from the feature amount, and selecting a plurality of songs stored in a music data storage unit in which the feature amount and the impression value are stored. A playlist generating step of generating a song list indicating a list composed of a plurality of songs, and a song constituting the song list A playlist replacement step of rearranging the order of the songs constituting the song list so that the first impression value list indicating the transition of the impression value is approximated to the first impression value list indicating the transition of the impression value of the transition information. And characterized in that:
[0022]
The music distribution system of the present invention is a music distribution system including a server that generates a playlist indicating the order of songs to be played by a playback device, and a terminal connected to the server via a network, The terminal includes a playback condition input unit for inputting a desired playlist length, and an image transition input unit for inputting transition information indicating a transition of an image of the desired playlist. A music source storage unit that stores music information, an impression value calculation unit that extracts a musical feature amount of a song, and calculates an impression value indicating an image when listening to music from the feature amount, and stores the feature amount and the impression value. Playlist data for generating a song list indicating a list composed of a plurality of songs by selecting the selected music data storage unit and the plurality of songs stored in the music data storage unit. And the first impression value list indicating the transition of the impression values of the songs constituting the song list, and the first impression value list indicating the transition of the impression values of the transition information. A playlist replacement unit for rearranging the order of songs and a music source acquisition unit for acquiring music sources of a plurality of songs described in the song list from a music source storage unit and outputting the music sources to a terminal; And a playback device that plays back the music source.
[0023]
BEST MODE FOR CARRYING OUT THE INVENTION
(Embodiment 1)
FIG. 1 is a block diagram showing a music playback device 100 according to Embodiment 1 of the present invention.
[0024]
The music source storage unit 11 is a storage device such as an HDD, and stores a plurality of music signals. A desired music signal data can be called up by a song ID such as a song name or a song number. The music signal is stored as compressed data such as waveform data by the linear PCM or AAC. These music sources can be added or deleted as appropriate, and are linked to a music data storage unit 31 described later.
[0025]
The impression value calculation unit 21 analyzes the physical characteristic amount of the input music signal, and calculates an impression value obtained by weighting and adding the characteristic amount. The physical feature amount includes a spectrum change degree (a degree of spectrum change between frames of a predetermined time length), an average number of pronunciations (sounding frequency of a sound pronounced in a music piece), and a sound non-periodicity (in a music piece). ), Beat period (time length equivalent to a quarter note of a song), beat cycle ratio (sound equivalent to a quarter note of a song, and sound that is predominantly pronounced) ), Beat intensity 1 (the intensity of a sound that is pronounced in a cycle corresponding to approximately half of the beat period), and pulse intensity 2 (a sound that is pronounced in a period that is approximately 1/4 of the beat period) , And a pulse intensity ratio (a ratio of pulse intensity 1 and pulse intensity 2) are analyzed. A predetermined weighting addition is performed on these feature amounts obtained by the analysis, and an impression value is calculated. The number of impression values is one or more, generally two to five. In the case of two, a two-dimensional space in which a large number of music pieces can be distinguished and distinguished by a combination of two impression value values. In the case of five, it becomes a five-dimensional space. A plurality of music pieces have respective coordinate values of impression values in this space.
[0026]
The impression value is an index indicating a feeling or emotion that a person feels when listening to music, such as intenseness, dynamism, exhilaration, simplicity, and softness. Each of these index values can be represented as a linear or non-linear function of the plurality of types of physical features. A five-dimensional space can be obtained by using five types of impression values of intenseness, dynamism, exhilaration, simplicity, and softness. Integrating the two of intenseness and dynamism into an active factor, and integrating the three of exhilaration, simplicity, and softness into an emotional factor, summing up two impression values of an active factor and an emotional factor. It becomes a two-dimensional space. Such a spatial coordinate system representing the impression value of the music is hereinafter referred to as an impression space. Since the impression value is a quantity representing the impression of listening to the music, if the impression values of a plurality of music are arranged in the impression space, they are arranged according to the similarity of the impressions. In other words, music pieces with similar impressions are laid out at close positions in the impression space. FIG. 3 shows an example of the impression space.
[0027]
An apparatus and method for registering data for music search and music search using such feature amounts and impression values are described in Japanese Patent Application No. 2001-082150. In the present invention, such data registration of feature amounts and impression values, music search, and an abbreviated name, an impression search method can be applied.
[0028]
The music data storage unit 31 stores the feature amount and the impression value calculated by the impression value calculation unit 21 corresponding to the music ID of the music in the music source storage unit 11.
[0029]
The playback condition input unit 41 receives the number of songs, time, or total storage capacity of a playlist desired by the user. The total storage capacity means the total amount of data of the music source that constitutes the playlist, and varies depending on the playing time of the music or the format of compressed audio.
[0030]
The image transition input unit 42 designates the transition of the image of the playlist desired by the user as the locus of the search center coordinate value when selecting a song using the impression space.
[0031]
As a method of specifying, there are a method of directly specifying by the user and a method of selecting a preset template. Also, the number of dimensions to be specified can be set independently for all dimensions of the impression space (Case 1), and can be set more easily by consolidating any number of dimensions into a smaller number of dimensions. Yes (Case 2).
[0032]
Here, a two-dimensional case of an active factor and an emotion factor will be described. At this time, the range in which the transition of the image is designated in the impression space (hereinafter referred to as a transition range) is within a circle having a center (0, 0) and a radius of 2. Here, “high-elevation” is defined as a new index integrating two-dimensional active factors and emotion factors. FIG. 4 is a diagram showing the correspondence between the impression space and the high lift. The lift of each song is (e) and the impression value (x, y) is
e = (− x + y) / √2 (1)
And −2 = <e = <2.
[0033]
[Case 1] When image transition is specified by two-dimensional impression section
FIG. 5 is a diagram illustrating a state in which a transition of an image is input to the impression space. In this way, a desired image transition can be specified by a straight line or a curve. At this time, not only can all be drawn by lines, but also a transition can be specified by specifying some relay points and interpolating between the relay points with a straight line.
[0034]
FIG. 6 is a diagram showing a template of transition of each factor. The transition of the image is plotted on the horizontal axis as time, and the vertical axis is plotted as a graph of the values of each dimension as a template. The desired image transition is determined by selecting a template for each dimension.
[0035]
[Case 2] Specifying image transition with high lift
FIG. 7 is a diagram showing a state in which a transition of an image is input to a graph of high lift. Using a graph in which the horizontal axis represents time and the vertical axis represents high lift, a change in high lift over time is input as a straight line or a curve. At this time, not only can all be drawn by lines, but also a transition can be specified by specifying some relay points and interpolating between the relay points with a straight line.
[0036]
FIG. 8 is a diagram showing a template of a high-elevation transition. A graph in which time is plotted on the abscissa and elevation is plotted on the ordinate is prepared as a template, and a desired image transition is determined by selecting a template.
[0037]
In the following, although the processing is different between [Case 1] and [Case 2], the transition of the image designated by the user in this way is collectively referred to as transition information.
[0038]
The image transition adjustment unit 50 includes a distribution analysis unit 51 and an image transition correction unit 52.
[0039]
The distribution analysis unit 51 analyzes the distribution of impression values of songs owned by the user. In the impression space, the absolute value of the maximum value and the minimum value of each of the active factor and the emotion factor becomes 2 or more if a large number of types of songs are registered. However, the songs registered by the user may be biased. Therefore, the distribution of the impression space is analyzed, and a playlist that matches the distribution of the user is generated.
[0040]
In a state where a large number of songs are registered, the transition range is defined as center coordinates (0, 0) and radius 2, whereas here, the center coordinates (dx, dy) of the user's individual transition range are defined. , Radius dr. At this time, the range (demin to demax) that the high lift takes is also automatically determined. The method will be described below.
[0041]
FIG. 9 is a diagram illustrating a method of determining a transition range of a user. For the impression values (x, y) of all songs, the maximum value and the minimum value of the X coordinate and the maximum value and the minimum value of the Y coordinate are searched. Let Xmax, Xmin, Ymax, and Ymin be respectively. Each of them is corrected so that the absolute value is within 2, that is, included in the transition range of the initial setting. At this time, the center coordinates (dx, dy) of the user's personal transition range are
(Dx, dy) = ((Xmax + Xmin) / 2, (Ymax + Ymin) / 2)) (2)
And Also, the radius dr at this time is
dr = min ((Xmax-Xmin) / 2, (Ymax-Ymin) / 2)) (3)
And Next, a line that passes through the center of the user's individual transition range and is parallel to y = −x is drawn, and a normal line is drawn from the intersection with the distribution range to y = −x, thereby obtaining the range of the user's individual high lift. Is determined (demin to demax). The data of the calculated transition range is sent to the image transition correction unit 52.
[0042]
The image transition correction unit 52 corrects the transition information of the image transition input unit 42 according to the transition range of the distribution analysis unit 51. The transition information of the image transition input unit 42 has a center (0, 0) and a radius of 2, and this is corrected using the result of the distribution analysis unit 51 as follows.
[0043]
[Case 1]
FIG. 10 is a diagram illustrating a method of correcting the transition information of the impression space. Each coordinate (x, y) in the transition information
nx = dr / 2 × x + dx (4)
ny = dr / 2 × y + dy (5)
(Nx, xy).
[0044]
[Case 2]
FIG. 11 is a diagram illustrating a method of correcting the high-elevation transition information. In response to the user's input (emin-emax), the high-elevation (e) in the transition information becomes the high-elevation range (demin-demax) with respect to the user's personal transition range.
ne = (demax-demin) / 4 × (e-emin) + demin (6)
Change to
[0045]
The music / voice determination unit 55 determines whether the music source registered in the music source storage unit 11 is music or voice. As the determination method, the impression value extracted by the impression value calculation unit 21 is used.
[0046]
The voice is particularly evident in the third and fourth quadrants away from the origin. Therefore, for the impression value (x, y) of each song,
<Condition 6> x ＾ 2 + y ＾ 2> (Th6) ＾ 2 (7)
<Condition 7> y <Th7 (8)
Those satisfying both are determined to be speech. In this way, the music ID of the music source determined to be audio is output to the impression value selection unit 61, and the corresponding music is not included in the playlist.
[0047]
The impression value search unit 61 generates a music list based on the transition information and the transition range received from the image transition correction unit 52 and the reproduction condition specified by the user received from the reproduction condition input unit 41. FIG. 12 is a diagram showing a music list. The song list describes a list number indicating the order of songs to be played and a song ID of the song played at that time.
[0048]
First, the reproduction condition designated by the user is quantified as follows and is set as C.
[0049]
<In the case of the number of songs> C = (the number of designated songs) (9)
<In case of time> C = (specified time) (10)
<In the case of total storage capacity> C = (specified total storage capacity) (11)
Then, a song is selected according to the transition information and added to the song list, and the process ends when the completion degree (V) of the playlist at a certain point reaches C or when there are no more selectable songs. . At this time, the point Vi to be added to the completion degree of the playlist when each song (Si) is selected is determined as follows.
[0050]
<In the case of the number of songs> Vi = 1 (12)
<Time> Vi = reproduction time of each song (13)
<Total storage capacity> Vi = storage capacity of each song... (14)
FIG. 13 is a flowchart showing a method of generating a music list. First, the total trajectory length L of the transition of the image is calculated from the transition of the impression space in [Case 1] and from the graph of the high lift in [Case 2]. Also, an index V = 0 representing the degree of completion of playlist generation is initialized (S102). Next, it is checked whether or not the music data storage unit 31 has a song with C <= V + Vi, that is, a song that can be added to the playlist (S103). If there is, the search center coordinate value is determined from the position of the trajectory length V / C * L in the transition information. FIG. 14 is a diagram illustrating a method of determining a search center coordinate value from an impression space. In [Case 1], the coordinate value of the position having the previous orbit length V / C * L is extracted from the total orbit length L, and the coordinates are directly used as the search center coordinate value. In the case of [Case 2], first, a lift is determined for determining a search center coordinate value, and the lift is converted into an impression value to determine a search center coordinate value. FIG. 15 is a diagram illustrating a method of determining a high lift of the search center coordinate value from a high lift graph. From the total orbit length L, the ellipticity at the position where the orbital length V / C * L is obtained is extracted, and the value is used as the ellipticity of the search center coordinate value. FIG. 16 is a diagram showing a method of converting a high lift into an impression value. A perpendicular line to y = −x is drawn from an intersection point lowered from y = −x from the high lift e to obtain an intersection line with the transition range of the user. One point on this intersection line is arbitrarily selected, and the coordinate value of this point is set as the search center coordinate value (S104). Next, in S103, a music piece having the closest impression value distance is selected from the search center coordinate values, and the music piece ID is added to the music list. However, the music determined to be voice by the music / voice determination unit 55 is not selected (S105). The degree of completion of playlist generation is updated as V = V + Vi (S106). If there is no song that satisfies the condition in S103, the generation of the playlist ends (S107). At this time, it is assumed that the number of songs constituting the playlist has become N.
[0051]
In S105, the tune is set to the shortest distance. However, it is possible to generate a playlist composed of different tunes every time by arbitrarily selecting from tunes that are close to some extent.
[0052]
The playlist replacement unit 71 rearranges the songs in the song list so as to match the transition information specified by the user. In the song list, when a song is selected, a song near the search center coordinate value is selected. Therefore, in some cases, even if "Emotion factor gradually increases", the order may be changed in the middle. . Thus, at first glance, it appears that the playlist is not correctly generated. Therefore, rearrangement is performed in the same order as the transition of the image of the transition information as follows.
[0053]
[Case 1]
For each song, the shortest distance point of the transition trajectory (if there are a plurality of shortest distance points, any one of them) is taken, and the music list is rearranged in the orbital order. At this time, the trajectories of the transition in the impression space may overlap two or more times. In this case, the songs are sorted in order for each overlapping trajectory.
[0054]
[Case 2]
First, according to the transition information, a high lift at a point where the number of reproduced music pieces is divided at equal intervals is obtained. The order of the lift of the p-th music is examined, and the result is put into the array oe. For example, if the number of played songs N = 4,
oe [1] = 2, oe [2] = 4, oe [3] = 1, oe [4] = 3 (15)
And so on. In this case, “the first song has the second highest lift”, “the second song has the fourth highest lift”, “the third song has the highest lift”, and “the fourth song has the third highest lift”. This means that the user desires "a song with a high degree of high lift". If there are a plurality of oe [x] with the same high lift, one of them is prioritized so as not to be in the same rank. Next, the lift of each song in the song list generated by the impression value search unit 61 is calculated, and the order of the song with the list number q is examined, and the result is entered into the array le. If there are a plurality of songs having the same high lift in the song list, one of them is prioritized so as not to be in the same rank. For example,
le [1] = 4, le [2] = 3, le [3] = 2, le [4] = 1 (16)
And so on. In this case, "the song of list number 1 has the fourth highest lift", "the song of list 2 has the third highest lift", "the song of list 3 has the second highest lift" , "The song of list number 4 has the highest lift". Then, the order of the songs having the r-th highest lift in the song list is changed so as to match the order of the high lift in the array oe. In this case, the music list is rearranged in the order of the list numbers “3 → 1 → 4 → 2”.
[0055]
Even if the order is not changed in this way, it is also possible for the impression value search unit 61 to select a tune in advance so as to match the transition information, for example, so that “the emotional factor gradually increases”. However, in this case, the algorithm becomes complicated, and in some cases, there is a possibility that there is no song to be selected at the end of the list. On the other hand, by changing the order of the music list once created by the playlist replacing unit 71, it is possible to reliably generate a playlist that matches the transition information specified by the user with a simple algorithm.
[0056]
The music source obtaining unit 81 obtains a music signal from the music source storage unit 11 based on the song ID, for a song to be included in the playlist for each list number of the song list. The acquired music signal is passed to the playback device 91.
[0057]
The reproducing device 91 has an amplifier and a speaker, and reproduces a music signal.
[0058]
(Embodiment 2)
FIG. 17 is a block diagram showing a music playback device 200 according to Embodiment 2 of the present invention. The music reproducing device 200 is different from the music reproducing device 100 shown in FIG. 1 in that the image transition adjusting unit 50 is not provided and the content of the playlist generating unit 60 is an impression value selecting unit 62. The other configuration is the same as that of the music reproducing device 100, and thus the description is omitted.
[0059]
The impression value selection unit 62 randomly selects a number of songs to be played from the songs registered in the music data storage unit 31, and generates a song list. Then, the music list is passed to the playlist replacing section 71.
[0060]
The playlist replacement unit 71 rearranges songs according to the transition information, as in the case of the music playback device 100. Thus, a playlist desired by the user can be generated only by rearranging the songs randomly selected by the impression value selection unit 62 as input by the image transition input unit 42.
[0061]
Since the music reproducing device 200 has a smaller configuration than the music reproducing device 100, it can perform processing with a small amount of calculation. This is effective when a playlist is generated in a device having only low-speed processing capability.
[0062]
The above-described playlist generation processing in Embodiments 1 and 2 of the present invention can be recorded on a recording medium in the form of a program. As the recording medium, any type of recording medium that can be read by a computer, such as a flexible disk or an optical disk, can be used. By installing the playlist generation processing program read from the recording medium in an arbitrary computer, the computer can function as a music playback device. The playlist generation processing may be executed by a playlist generation processing device built in or connected to the computer, or the computer may execute at least a part of the playlist generation processing on software.
[0063]
FIG. 18 illustrates one embodiment of a computer that executes a playlist generation process. Computer 300 includes a disk drive device 301 that reads a program from recording medium 310 for storing a program for causing computer 300 to execute a playlist generation process, an operation unit 302 that executes a playlist generation process, and a music source storage. It includes a storage unit 303 functioning as the unit 11 and the music data storage unit 31, a reproduction condition input unit 41, an image transition input unit 42, a reproduction device 91, and a bus 304. The calculation unit 302 includes a CPU, a memory, and the like (not shown), and includes the impression value calculation unit 21, the image transition adjustment unit 50, the music / sound determination unit 55, the playlist generation unit 60, and the playlist illustrated in FIGS. The same processing as the replacement unit 71 is performed.
[0064]
The program may be provided by a recording medium, or may be provided by data distribution over the Internet or the like.
[0065]
(Embodiment 3)
FIG. 19 is a diagram showing a music distribution system 400 according to Embodiment 3 of the present invention.
[0066]
The music distribution system 400 includes a network 410, a server 420, and a terminal 430 such as a personal computer or a music server. In the music distribution system 400, when the content of the playlist desired by the terminal 430 is input to the reproduction condition input unit 41 and the image transition input unit 42, a playlist suitable for the input condition is automatically generated on the server 420 side. To the terminal 430.
[0067]
The network 410 is connected to a server 420 that distributes music and a terminal 430 that receives the distributed music. A plurality of terminals 430-1 to 430-n (n is a natural number) having the same function as the terminal 430 are connected to the network 410.
[0068]
The music distribution system 400 divides the components of the music playback devices 100 and 200 according to the first and second embodiments and arranges them on a network. Thus, even when the user does not have the music source storage unit 11 at hand, it is possible to download a playlist suitable for the image he or she wants from the server 420 via the network 410.
[0069]
The server 420 includes a music source storage unit 11, an impression value calculation unit 21, a music data storage unit 31, a music / sound determination unit 55, a playlist generation unit 60, a playlist replacement unit 71, and a music source acquisition unit 81. Is provided. The server 420 may include the image transition adjusting unit 50 of the music reproducing apparatus 100. However, since the large-scale music database usually stores a sufficient variety of music sources, the image transition adjusting unit 50 is omitted. May be done. The terminal 430 includes a playback condition input unit 41, an image transition input unit 42, a recording medium 431 for storing a playlist distributed from the server 420, and a playback device 91. The recording medium 431 is a personal computer, a hard disk of a music server, or the like, and stores a playlist distributed from the server. The music included in the playlist is input to the playback device 91 and played.
[0070]
According to the music distribution system 400 of the present embodiment, when a user inputs a condition to the terminal 430 connected to the network 410, the server 420 generates a playlist suitable for the condition and distributes the playlist to the terminal 430. Can be realized.
[0071]
【The invention's effect】
By expressing the image of the music by the transition of the impression value, it is possible to generate a reproduction music list including the music without registering new knowledge even when registering a new music. In addition, since the user's preference can be reflected in the process of generating the reproduction music list from the impression value list, the user's satisfaction can be further improved.
[0072]
In the present invention, the transition of the playlist desired by the user is corrected according to the distribution of impression values of the songs owned by the user. As a result, it is possible to generate a playlist that matches a song owned by the user.
[0073]
In addition, in the present invention, whether a music or a sound is determined using an impression value of a registered music and a sound signal is not included in the playlist, a highly satisfactory play composed only of music is performed. Lists can be generated.
[0074]
Further, in the present invention, the transition of the impression value of the song forming the playlist is rearranged so as to be similar to the transition of the impression value of the transition information designated by the user, thereby satisfying the user's desire. Can be generated.
[Brief description of the drawings]
FIG. 1 is a block diagram showing a music playback device according to an embodiment of the present invention.
FIG. 2 is a diagram showing a conventional case where the order of music images is changed.
FIG. 3 is a diagram showing an impression space according to the embodiment of the present invention.
FIG. 4 is a diagram showing a correspondence between an impression space and a high lift according to the embodiment of the present invention;
FIG. 5 is a diagram showing a state of inputting a transition of an image into an impression space according to the embodiment of the present invention;
FIG. 6 is a diagram showing a template of transition of each factor according to the embodiment of the present invention.
FIG. 7 is a diagram showing a state of inputting a transition of an image into a graph of high lift according to the embodiment of the present invention.
FIG. 8 is a diagram showing a template of a high-elevation transition according to the embodiment of the present invention;
FIG. 9 is a diagram showing a method for determining a transition range of a user according to the embodiment of the present invention;
FIG. 10 is a diagram showing a method for correcting transition information of an impression space according to the embodiment of the present invention.
FIG. 11 is a diagram showing a method of correcting high-elevation transition information according to the embodiment of the present invention.
FIG. 12 is a diagram showing a music list according to the embodiment of the present invention.
FIG. 13 is a flowchart showing a music list generation method according to the embodiment of the present invention;
FIG. 14 is a diagram showing a method of determining a search center coordinate value from an impression space according to the embodiment of the present invention.
FIG. 15 is a diagram showing a method of determining a lift of a search center coordinate value from a graph of the lift according to the embodiment of the present invention.
FIG. 16 is a diagram showing a method of converting a high lift into an impression value according to the embodiment of the present invention.
FIG. 17 is a diagram showing a music playback device according to an embodiment of the present invention.
FIG. 18 is a diagram illustrating a computer that executes a playlist generation process according to the embodiment of this invention.
FIG. 19 is a diagram showing a music distribution system according to an embodiment of the present invention.
[Explanation of symbols]
11 Music source storage
21 Impression value calculator
31 Music data storage
41 Playback condition input section
42 Image transition input section
50 Image transition adjustment unit
51 Distribution analysis unit
52 Image transition correction unit
55 Music / Speech Judgment Unit
60 Playlist generation unit
61 Impression value search section
62 Impression value selection section
71 Playlist Replacement Unit
81 Music Source Acquisition Department
91 Playback device
100 Music playback device
200 Music playback device
300 computer
301 Disk drive device
302 arithmetic unit
303 storage unit
304 bus
310 recording medium
400 music distribution system
410 Network
420 server
430, 430-1 to 420-n terminal
431 Recording medium

Claims (13)

A playlist generation device that generates a playlist indicating the order of songs to be played by the playback device,
A playback condition input unit for inputting a desired playlist length;
An image transition input unit for inputting transition information indicating a transition of a desired playlist image,
A music source storage unit in which a plurality of songs are stored;
An impression value calculation unit that extracts a musical feature amount of the song and calculates an impression value indicating an image when listening to music from the feature amount;
A music data storage unit in which the feature amount and the impression value are stored;
A playlist generation unit that generates a song list indicating a list composed of the plurality of songs by selecting a plurality of songs stored in the music data storage unit;
The song list is configured such that a first impression value list indicating a transition of an impression value of a song constituting the song list is similar to a second impression value list indicating a transition of an impression value of the transition information. A playlist generation device, comprising: a playlist replacement unit that rearranges the order of songs. The playlist generation unit,
Using the playback conditions and the transition information, a third impression value list indicating transitions of impression values is generated, and the plurality of songs that give transitions in impression values that are close to the third impression value list are generated by the music. The play list generation device according to claim 1, further comprising an impression value search unit that selects from the data storage unit. The play list generation device according to claim 1, further comprising a music / speech determination unit that determines whether the song is music or voice using the impression value. A distribution analysis unit that calculates a transition range indicating a distribution of impression values of songs registered in the music data storage unit;
The playlist generation device according to claim 1, further comprising: an image transition correction unit that corrects the transition information according to the transition range. A playlist generation device according to any one of claims 1 to 4,
A music source acquisition unit that acquires music sources of a plurality of songs described in the song list from the music source storage unit;
A music playback device, comprising: the playback device that plays a music source. A playlist generating method for generating a playlist indicating an order of a song to be reproduced by a reproducing device,
A reproduction condition input step of inputting a desired playlist length;
An image transition input step of inputting transition information indicating a transition of a desired playlist image,
An impression value calculating step of extracting a musical feature amount of the song and calculating an impression value indicating an image when listening to music from the feature amount;
A playlist generation step of generating a song list indicating a list composed of the plurality of songs by selecting a plurality of songs stored in the music data storage unit in which the feature amount and the impression value are stored;
The song list is configured such that a first impression value list indicating a transition of an impression value of a song constituting the song list is similar to a second impression value list indicating a transition of an impression value of the transition information. A playlist replacement method for rearranging the order of songs. The playlist generation step includes:
A third impression value list indicating a transition of an impression value is generated using the reproduction condition and the transition information. 7. The playlist generation method according to claim 6, further comprising an impression value search step of selecting from a music data storage unit in which a value is stored. The play list generation method according to claim 6, further comprising a music / speech determination step of determining whether the song is music or voice using the impression value. A distribution analysis step of calculating a transition range indicating a distribution of impression values of songs registered in the music data storage unit in which the impression values are stored;
7. The playlist generation method according to claim 6, further comprising an image transition modification step of modifying the transition information according to the transition range. Music source acquisition step of acquiring music sources of a plurality of songs described in the song list from a music source storage unit;
The play list generation method according to any one of claims 6 to 9, further comprising a reproduction step of reproducing the music source. A program for causing a computer to execute a playlist generation process for generating a playlist indicating the order of songs to be played by the playback device,
The generation processing includes:
A reproduction condition input step of inputting a desired playlist length;
An image transition input step of inputting transition information indicating a transition of a desired playlist image,
An impression value calculating step of extracting a musical feature amount of the song and calculating an impression value indicating an image when listening to music from the feature amount;
A playlist generation step of generating a song list indicating a list composed of the plurality of songs by selecting a plurality of songs stored in the music data storage unit in which the feature amount and the impression value are stored;
The song list is configured such that a first impression value list indicating a transition of an impression value of a song constituting the song list is similar to a second impression value list indicating a transition of an impression value of the transition information. A playlist replacement step of rearranging the order of songs. A computer-readable recording medium that records a program for causing a computer to execute a playlist generation process that generates a playlist indicating the order of songs to be played by a playback device,
The generation processing includes:
A reproduction condition input step of inputting a desired playlist length;
An image transition input step of inputting transition information indicating a transition of a desired playlist image,
An impression value calculating step of extracting a musical feature amount of the song and calculating an impression value indicating an image when listening to music from the feature amount;
A playlist generation step of generating a song list indicating a list composed of the plurality of songs by selecting a plurality of songs stored in the music data storage unit in which the feature amount and the impression value are stored;
The song list is configured such that a first impression value list indicating a transition of an impression value of a song constituting the song list is similar to a second impression value list indicating a transition of an impression value of the transition information. A playlist replacement step of rearranging the order of songs. A server that generates a playlist indicating the order of songs to be played by the playback device;
A terminal connected to the server via a network,
Music distribution system comprising
The terminal is
A playback condition input unit for inputting a desired playlist length;
An image transition input unit for inputting transition information indicating transition of a desired playlist image,
The server comprises:
A music source storage unit in which a plurality of songs are stored;
An impression value calculation unit that extracts a musical feature amount of the song and calculates an impression value indicating an image when listening to music from the feature amount;
A music data storage unit in which the feature amount and the impression value are stored;
A playlist generation unit that generates a song list indicating a list composed of the plurality of songs by selecting a plurality of songs stored in the music data storage unit;
The song list is configured such that a first impression value list indicating a transition of an impression value of a song constituting the song list is similar to a second impression value list indicating a transition of an impression value of the transition information. A playlist replacement unit that rearranges the order of songs, and a music source acquisition unit that acquires music sources of a plurality of songs described in the song list from the music source storage unit and outputs the music sources to the terminal.
The terminal is
A storage unit for storing the plurality of music sources received from the server,
A music distribution system, comprising: the reproduction device that reproduces a music source.

Images