Granted Patent B2

US 10,242,097 · App. 14/671,979 · Granted Mar 26, 2019

Inventors: Jacquelyn Fuzell-Casey (Mercer Island, WA); Skyler Fuzell-Casey (Portland, OR); Donald Ryan (Auburn, AL); Timothy D. Casey (Mercer Island, WA)

Assignee: APERTURE INVESTMENTS, LLC

G06F17/30743G06Q10/00

Abstract

Claims (72)

1. A method for assigning mood to music, comprising:

determining a first score corresponding to a rhythm of the music, wherein determining the first score includes:

determining an average of beats per minute for the music;

determining an average of time signatures for the music; and

mapping the average of beats per minute and the average of time signatures to the first score;

determining a second score corresponding to a texture of the music;

determining a third score corresponding to a pitch of the music, wherein the first score, the second score and the third score define a rhythm, texture and pitch (RTP) vector for the music;

comparing the RTP vector to a set of predetermined RTP vectors, wherein each predetermined RTP vector among the set of predetermined RTP vectors is unique and corresponds to only one predetermined mood among a set of predetermined moods; and

assigning a predetermined mood among the set of predetermined moods to the music based on a match between the RTP vector and one of the predetermined RTP vectors among the set of predetermined RTP vectors.

2. The method as recited in claim 1 , wherein determining the average of beats per minute for the music and determining the average of time signatures for the music includes:

generating a static representation of the music based on frequency data sampled from the content;

determining beats per minute for the music from the static representation;

determining time signatures for the music from the static representation;

averaging the beats per minute; and

averaging the time signatures.

3. A method for assigning mood to music, comprising:

determining a first score corresponding to a rhythm of the music;

determining a second score corresponding to a texture of the music, wherein determining the second score includes:

determining an average of frequency densities for the music; and

mapping the average of frequency densities to the second score;

determining a third score corresponding to a pitch of the music, wherein the first score, the second score and the third score define a rhythm, texture and pitch (RTP) vector for the music;

comparing the RTP vector to a set of predetermined RTP vectors, wherein each predetermined RTP vector among the set of predetermined RTP vectors is unique and corresponds to only one predetermined mood among a set of predetermined moods; and

assigning a predetermined mood among the set of predetermined moods to the music based on a match between the RTP vector and one of the predetermined RTP vectors among the set of predetermined RTP vectors.

4. The method as recited in claim 3 , wherein determining the average of frequency densities includes;

generating a static representation of the music based on frequency data sampled from the music; and

averaging the density of frequency data within the static representation.

5. A method for assigning mood to music, comprising:

determining a first score corresponding to a rhythm of the music;

determining a second score corresponding to a texture of the music;

determining a third score corresponding to a pitch of the music, wherein determining a third score includes:

determining a chromagram for the music;

determining a chromagram vector for the music from the chromagram;

comparing the chromagram vector to a set of predetermined chromagram vectors, wherein each predetermined chromagram vector among the set of predetermined chromagram vectors is unique and corresponds to one value of the third score among a set of predetermined values for the third score; and

assigning a value of the third score to the music based on a match between the chromagram vector and one of the predetermined chromagram vectors among the set of predetermined chromagram vectors;

wherein the first score, the second score and the third score define a rhythm, texture and pitch (RTP) vector for the music;

comparing the RTP vector to a set of predetermined RTP vectors, wherein each predetermined RTP vector among the set of predetermined RTP vectors is unique and corresponds to only one predetermined mood among a set of predetermined moods; and

assigning a predetermined mood among the set of predetermined moods to the music based on a match between the RTP vector and one of the predetermined RTP vectors among the set of predetermined RTP vectors.

6. The method as recited in claim 5 , wherein determining the chromagram for the music includes:

generating a static representation of the content based on frequency data sampled from the content; and

filtering the static representation to capture intensity differences represented in the static representation and generating a filtered representation of the music as the chromagram.

7. A method for assigning mood to music, comprising:

determining a first score corresponding to a rhythm of the music wherein the rhythm includes time signatures of the music;

determining a second score corresponding to a texture of the music, wherein the texture includes frequency densities for the music;

determining a third score corresponding to a pitch of the music, wherein the first score, the second score and the third score define a rhythm, texture and pitch (RTP) vector for the music;

comparing the RTP vector to a set of predetermined RTP vectors, wherein each predetermined RTP vector among the set of predetermined RTP vectors is unique and corresponds to only one predetermined mood among a set of predetermined moods; and

assigning a predetermined mood among the set of predetermined moods to the music based on a match between the RTP vector and one of the predetermined RTP vectors among the set of predetermined RTP vectors.

8. The method as recited in claim 7 , wherein the first score, the second score and the third score are integers between 1 and 5.

9. The method as recited in claim 8 , wherein the set of predetermined moods consists of happy, excited, manic, cautious, peaceful and sad.

10. The method as recited in claim 8 , wherein the set of predetermined moods includes happy, excited, manic, cautious, peaceful and sad.

11. The method as recited in claim 10 , wherein the one predetermined mood to which each predetermined RTP vector is assigned is determined by a classifier trained with a subset of predetermined RTP vectors among the set of predetermined RTP vectors.

12. The method as recited in claim 11 , wherein the classifier is a multiclass support vector machine.

13. The method as recited in claim 11 , wherein the classifier is trained with at least predetermined RTP vectors of (3,4,4), (3,3,5), (3,5,5), and (3,3,4) for happy.

14. The method as recited in claim 11 , wherein the classifier is trained with at least predetermined RTP vectors of (2,4,4), (3,4,3), (3,5,3), and (2,2,4) for excited.

15. The method as recited in claim 11 , wherein the classifier is trained with at least predetermined RTP vectors of (5,4,3), (4,4,2), (4,5,3), (4,4,3), and (4,5,2) for manic.

16. The method as recited in claim 11 , wherein the classifier is trained with at least predetermined RTP vectors of (2,4,2), (3,4,2), (3,3,2), (3,3,3) and (2,3,3) for cautious.

17. The method as recited in claim 11 , wherein the classifier is trained with at least predetermined RTP vectors of (3,2,3), (2,2,3), (3,2,2), (4,2,1), and (1,1,3) for peaceful.

18. The method as recited in claim 11 , wherein the classifier is trained with at least predetermined RTP vectors of (1,2,1), (2,3,1), (1,1,2), (3,5,2), and (2,2,2) for sad.

19. The method as recited in claim 7 , wherein the music is a song, and further comprising:

repeating each step for each song among a plurality of songs until each song has been assigned to predetermined moods among the set of predetermined moods; and

placing all of the songs within at least one predetermined mood among the set of predetermined moods on a playlist.

20. The method as recited in claim 7 , wherein the music is a song, and further comprising:

repeating each step for each song among a plurality of songs until each song has been assigned to predetermined moods among the set of predetermined moods;

providing a user with one or more preferences based on predetermined attributes of each song among the plurality of songs;

enabling the user to create a custom playlist from the songs assigned to a single predetermined mood by:

placing all of the songs assigned to the single predetermined mood on a playlist;

enabling the user to select at least one preference from the one or more preferences in order to remove songs from the playlist.

21. The method as recited in claim 20 , wherein the one or more preferences are derived from metadata associated with one or more of the songs among the plurality of songs.

22. A content selection and organization system, comprising:

a sampler configured to sample frequency data from a plurality of content and to generate rhythm data, texture data, and static representations based on the sampled frequency data, wherein the content includes music;

a filter configured to capture intensity differences represented in the static representations and generate pitch data;

an analyzer configured to generate a first score based on the rhythm data, wherein the rhythm data includes time signatures of the music, a second score based on the texture data, wherein the texture data includes frequency densities for the music, and a third score based on the pitch data, wherein the first score, the second score and the third score define a rhythm, texture and pitch (RTP) vector for the music, wherein the analyzer is further configured to compare the RTP vector to a set of predetermined RTP vectors, wherein each predetermined RTP vector among the set of predetermined RTP vectors is unique and corresponds to only one predetermined mood among a set of predetermined moods; and wherein the analyzer is further configured to assign a predetermined mood among the set of predetermined moods to the content based on a match between the RTP vector and one of the predetermined RTP vectors among the set of predetermined RTP vectors; and

a user interface configured to enable a user to view and to select content from the plurality of content based on the assigned predetermined moods.

Assignments (3)

Continuity (8)