Music selection and organization using audio fingerprints

• US 9,875,304 B2
• Filed: 03/27/2015
• Issued: 01/23/2018
• Est. Priority Date: 03/14/2013
• Status: Active Grant

• Current Assignee: Muvox LLC
• Original Assignee: Aperture Investments LLC
• Inventors: Fuzell-Casey, Jacquelyn, Fuzell-Casey, Skyler, Casey, Timothy D.
• Primary Examiner(s): Gilliard, Delomia L
• Application Number:
  US14/671,973
• Publication Number:
  US 20150205864A1
• Time in Patent Office: 1,033 Days
• Field of Search: None
• CPC Class Codes: G06F 16/683 using metadata automatically derived from the content
• CPC Class Codes: G06Q 10/00 Administration; Management
• CPC Class Codes: G06Q 30/0631 Item recommendations
• CPC Class Codes: G06T 17/20 Finite element generation, e.g. wire-frame surface description, tesselation
• CPC Class Codes: G10L 19/018 Audio watermarking, i.e. embedding inaudible data in the audio signal

Abstract

A content selection system and method for identifying and organizing moods in content using objectively measured scores for rhythm, texture and pitch (RTP) and clustered into six mood classifications based on an objective analysis of the measured scores. Digitized representations of the content may also be identified and organized based on the content'"'"'s frequency data, three-dimensional shapes derived from the digitized representations, and colors derived from the frequency data. Each piece of content may be identified by at least a mood shape, but may also be identified by a mood color and/or a mood based on the clustered RTP scores and/or the digitized representation. Users of the selection system may be able to view the moods identified in the different manners, or combinations of two or three mood identifying manners and select and organize content based on the identified moods.

Claims

1. A method for identifying mood in content, comprising:

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

filtering the static visual representation to capture intensity differences represented in the static visual representation and generating a filtered representation of the content;

encoding the filtered representation to create a digitized representation of the content;

identifying characteristics within the digitized representation that correspond to a first set of one or more moods among a set of moods; and

generating a three-dimensional shape that corresponds to the first set of one or more moods.

2. The method for identifying mood in content as recited in claim 1, wherein the three-dimensional shape is a wire frame.

3. The method for identifying mood in content as recited in claim 1, wherein the three-dimensional shape is a solid.

4. The method for identifying mood in content as recited in claim 1, wherein the three-dimensional shape includes at least one angle and one side and wherein a degree of the angle and a length of the side identify a percentage of a mood among the first set of one or more moods identified in the content.

5. The method for identifying mood in content as recited in claim 1, further comprising:

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

identifying predominant colors in the color representation and deleting less predominant colors from the color representation to generate a set of one or more colors representing one or more moods in the content; and

associating a second set of one or more moods to the set of one or more colors representing the content.

6. The method for identifying mood in content as recited in claim 5, wherein generating a color representation of the content based on the frequency data includes determining tristimulus values based on the frequency data and generating the color representation based on the tristimulus values.

7. The method for identifying mood in content as recited in claim 6, wherein the tristimulus values include a first tristimulus value based on a relative weight of a 1st harmonic frequency in the content, a second tristimulus value based on a relative weight of the 2nd, 3rd, and 4th harmonics in the content, and a third tristimulus value based on a relative weight of all remaining harmonics in the content.

8. The method for identifying mood in content as recited in claim 5, wherein generating a color representation of the content based on the frequency data includes determining a melody, a harmony and a rhythm in the content and generating the color representation based on a combination of the melody, the harmony and the rhythm.

9. The method for identifying mood in content as recited in claim 1, further comprising identifying a third set of one or more moods among the set of moods based on the digitized representation.

21. The method for identifying mood in content as recited in claim 1, wherein the static visual representation is one or more of a spectrogram, a chromagram or an audio fingerprint.

10. A content selection system, comprising:

a sampler configured to sample frequency data from a plurality of content and to generate static visual representations based on the sampled frequency data;

a filter configured to capture intensity differences represented in the static visual representations and to generate filtered representations of the plurality of content;

an encoder configured to encode the filtered representations and to create digitized representations of the plurality of content based on the encoded filtered representations;

an analyzer configured to identify characteristics within the digitized representations that correspond to a plurality of moods and to generate three-dimensional shapes that correspond to the plurality of moods; and

a user interface configured to enable a user to view the three-dimensional shapes and to select content from the plurality of content based on the three-dimensional shapes.

11. The content selection system as recited in claim 10, wherein the three-dimensional shapes are based on wire frames.

12. The content selection system as recited in claim 10, wherein the three-dimensional shapes are based on solids.

13. The content selection system as recited in claim 10, wherein each of the three-dimensional shapes includes at least one angle and one side and wherein a degree of the angle and a length of the side identify a percentage of a mood among the plurality of moods.

14. The content selection system as recited in claim 10, wherein the analyzer is further configured to generate color representations of the plurality of content based on the frequency data sampled from the content, to identify predominant colors in the color representations and to delete less predominant colors from the color representations to generate a plurality of colors representing moods in the content, and to associate the plurality of moods representing the content based on the plurality of colors.

15. The content selection system as recited in claim 14, wherein the color representations of the plurality of content are based on tristimulus values based on the frequency data.

16. The content selection system as recited in claim 15, wherein the tristimulus values include a first tristimulus value based on a relative weight of a 1st harmonic frequency in the plurality of content, a second tristimulus value based on a relative weight of the 2nd, 3rd, and 4th harmonics in the plurality of content, and a third tristimulus value based on a relative weight of all remaining harmonics in the plurality of content.

17. The content selection system as recited in claim 14, wherein the color representations of the plurality of content are based combinations of melodies, harmonies and rhythms in the plurality of content.

18. The content selection system as recited in claim 10, wherein the plurality of content includes music.

19. The content selection system as recited in claim 10, wherein the plurality of content includes video that includes music.

20. The content selection system as recited in claim 10, wherein the analyzer is further configured to identify the plurality of moods from the digitized representations.

22. The content selection system as recited in claim 10, wherein the static visual representation is one or more of a spectrogram, a chromagram or an audio fingerprint.