Media rights management using melody identification
First Claim
1. A computer-implemented method for determining ownership rights to a composition, the method comprising:
- receiving, by a computing device, an input media file, the input media file comprising an audio clip digitally representing a sound recording;
processing, by the computing device, the audio clip to extract dominant musical tones in the audio clip at a plurality of times within the audio clip, the extracted tones corresponding to notes of a musical scale;
determining, by the computing device, pitch intervals between consecutive ones of the dominant musical tones extracted from the audio clip;
generating, by the computing device, an input melody fingerprint from a sequence of the pitch intervals for the audio clip, the input melody fingerprint representing a melody of a composition that is embodied in the audio clip, the melody fingerprint being invariant to parameters used to perform or record a performance of the composition;
querying, by the computing device, a reference database for a reference melody fingerprint that matches the input melody fingerprint, each reference melody fingerprint in the reference database representing a melody that is embodied in a reference sound recording;
responsive to a matching reference melody fingerprint for the input melody fingerprint being found, determining by the computing device, ownership information for the composition from ownership information that is associated with the matching reference melody fingerprint; and
generating, by the computing device, a claim on behalf of an owner of the composition that matches the composition, the owner determined from the ownership information.
2 Assignments
0 Petitions
Accused Products
Abstract
A content recognition system operates in conjunction with a media hosting service to identify hosted media content and ownership rights associated with the hosted content. By applying melody recognition, the content recognition system can identify compositions embodied in hosted media content even when these compositions do not precisely match any known sound recording. Thus, the content recognition system is beneficially able to detect, for example, recorded cover performances and recorded live performances embodied in hosted media content. Once identified, ownership information is determined and the media hosting service can carry out appropriate rights management policies associated with the content such as monetizing or blocking the protected content.
38 Citations
34 Claims
-
1. A computer-implemented method for determining ownership rights to a composition, the method comprising:
-
receiving, by a computing device, an input media file, the input media file comprising an audio clip digitally representing a sound recording; processing, by the computing device, the audio clip to extract dominant musical tones in the audio clip at a plurality of times within the audio clip, the extracted tones corresponding to notes of a musical scale; determining, by the computing device, pitch intervals between consecutive ones of the dominant musical tones extracted from the audio clip; generating, by the computing device, an input melody fingerprint from a sequence of the pitch intervals for the audio clip, the input melody fingerprint representing a melody of a composition that is embodied in the audio clip, the melody fingerprint being invariant to parameters used to perform or record a performance of the composition; querying, by the computing device, a reference database for a reference melody fingerprint that matches the input melody fingerprint, each reference melody fingerprint in the reference database representing a melody that is embodied in a reference sound recording; responsive to a matching reference melody fingerprint for the input melody fingerprint being found, determining by the computing device, ownership information for the composition from ownership information that is associated with the matching reference melody fingerprint; and generating, by the computing device, a claim on behalf of an owner of the composition that matches the composition, the owner determined from the ownership information. - View Dependent Claims (2, 3, 4, 5, 6, 7, 8, 9)
-
-
10. A computer-implemented method for determining ownership rights to a sound recording, comprising:
-
receiving, by a computing device, an input media file, the input media file comprising an audio clip; generating, by the computing device, an input audio fingerprint for the audio clip, the audio fingerprint representing an input sound recording of a composition that is embodied in the audio clip; querying, by the computing device, a reference audio fingerprint database for a reference audio fingerprint that matches the input audio fingerprint, each reference audio fingerprint in the reference audio fingerprint database representing a reference sound recording; and responsive to a matching reference audio fingerprint for the input audio fingerprint in the reference database being found, determining, by the computing device, ownership information for the input sound recording from ownership information that is associated with the matching reference audio fingerprint; responsive to a matching reference audio fingerprint for the input audio fingerprint not being found; generating, by the computing device, an input melody fingerprint for the audio clip, the input melody fingerprint different than the input audio fingerprint, and the input melody fingerprint representing a melody of the composition, the melody fingerprint being invariant to parameters used to perform or record a performance of the composition; querying, by the computing device, a reference melody fingerprint database for a reference melody fingerprint that matches the input melody fingerprint, each reference melody fingerprint in the reference melody fingerprint database representing a melody that is embodied in a reference sound recording; and responsive to a matching reference melody fingerprint for the input melody fingerprint being found, determining, by the computing device, ownership information for the composition from ownership information that is associated with the matching reference melody fingerprint. - View Dependent Claims (11, 12)
-
-
13. A computer-implemented method for generating a searchable reference database of ownership rights for audio compositions, the method comprising:
-
receiving, by a computing device, a media file digitally representing a reference sound recording embodying a composition and receiving composition ownership information indicating one or more entities having ownership rights to the composition; processing, by the computing device, the media file to extract dominant musical tones in the reference sound recording at a plurality of times within the reference sound recording, the extracted tones corresponding to notes of a musical scale; determining, by the computing device, pitch intervals between consecutive ones of the dominant musical tones extracted from the reference sound recording; generating, by the computing device, a melody fingerprint from a sequence of the pitch intervals for the reference sound recording, the melody fingerprint representing a melody of the composition embodied in the reference sound recording in a manner that is invariant to parameters used to perform or record the performance of the composition; and storing to a memory, the melody fingerprint and the associated composition ownership information in the searchable reference database. - View Dependent Claims (14, 15, 16, 17)
-
-
18. A system for determining ownership rights to a composition, the system comprising:
-
a processor; and a non-transitory computer-readable storage medium coupled to the processor, wherein the non-transitory computer-readable storage medium stores instructions that when executed cause the processor to; receive an input media file, the input media file comprising an audio clip digitally representing a sound recording; extract dominant musical tones in the audio clip at a plurality of times within the audio clip, the extracted tones corresponding to notes of a musical scale; determine pitch intervals between consecutive ones of the dominant musical tones extracted from the audio clip; generate an input melody fingerprint from a sequence of the pitch intervals for the audio clip, the input melody fingerprint representing a melody of a composition that is embodied in the audio clip, the melody fingerprint being invariant to parameters used to perform or record a performance of the composition; query a reference database for a reference melody fingerprint that matches the input melody fingerprint, each reference melody fingerprint in the reference database representing a melody that is embodied in a reference sound recording; responsive to a matching reference melody fingerprint for the input melody fingerprint being found, determine ownership information for the composition from ownership information that is associated with the matching reference melody fingerprint; generate a claim on behalf of an owner of the composition that matches the composition, the owner determined from the ownership information. - View Dependent Claims (19, 20, 21, 22, 23, 24, 25, 26)
-
-
27. A system for determining ownership rights to a sound recording, the system comprising:
-
a processor; and a non-transitory computer-readable storage medium coupled to the processor, wherein the non-transitory computer-readable storage medium stores instructions that when executed cause the processor to; receive an input media file, the input media file comprising an audio clip; generating an input audio fingerprint for the audio clip, the audio fingerprint representing an input sound recording of a composition that is embodied in the audio clip; query a reference audio fingerprint database for a reference audio fingerprint that matches the input audio fingerprint, each reference audio fingerprint in the reference audio fingerprint database representing a reference sound recording; and responsive to a matching reference audio fingerprint for the input audio fingerprint in the reference database being found, determine ownership information for the input sound recording from ownership information that is associated with the matching reference audio fingerprint; responsive to a matching reference audio fingerprint for the input audio fingerprint not being found; generate an input melody fingerprint for the audio clip, the input melody fingerprint different than the input audio fingerprint, and the input melody fingerprint representing a melody of the composition, the melody fingerprint being invariant to parameters used to perform or record a performance of the composition; query a reference melody fingerprint database for a reference melody fingerprint that matches the input melody fingerprint, each reference melody fingerprint in the reference melody fingerprint database representing a melody that is embodied in a reference sound recording; and responsive to a matching reference melody fingerprint for the input melody fingerprint is found, determine ownership information for the composition from ownership information that is associated with the matching reference melody fingerprint. - View Dependent Claims (28, 29)
-
-
30. A system for generating a searchable reference database of ownership rights for audio compositions, the system comprising:
-
a processor; and a non-transitory computer-readable storage medium coupled to the processor, wherein the non-transitory computer-readable storage medium stores instructions that when executed cause the processor to; receive a media file digitally representing a reference sound recording embodying a composition and composition ownership information indicating one or more entities having ownership rights to the composition; process the media file to extract dominant musical tones in the reference sound recording at a plurality of times within the reference sound recording, the extracted tones corresponding to notes of a musical scale; determine pitch intervals between consecutive ones of the dominant musical tones extracted from the reference sound recording; generate a melody fingerprint from a sequence of the pitch intervals for the reference sound recording, the melody fingerprint representing a melody of the composition embodied in the reference sound recording in a manner that is invariant to parameters used to perform or record the performance of the composition; and store the melody fingerprint and the associated composition ownership information in the searchable reference database. - View Dependent Claims (31, 32, 33, 34)
-
Specification