Detecting repeating content in broadcast media
First Claim
Patent Images
1. A method, comprising:
- determining, by a system including a processor, a first match Mh between a first audio descriptor representing a first recording at a first time step in an environment and a first reference descriptor, the first match Mh having-a first confidence score Ch indicative of a confidence of the first match, the first time step having a time step length l;
determining, by the system, a second match M0 between a second audio descriptor representing a second recording at a second time step in the environment and a second reference descriptor, the second match M0 having a second confidence score C0 indicative of a confidence of the second match, the second time step having the time step length l, wherein the first time step is prior to the second time step, and the first match Mh and the second match M0 are non-identity matches determined using a direct or locality sensitive hashing function and a validation process to select a most accurate match out of a plurality of candidate matches, wherein the first confidence score Ch and the second confidence score C0 are based upon a log-likelihood function given by an audio fingerprinting process;
discounting, by the system, the first confidence score Ch by a discount value l/L to generate a discounted first confidence score Ch−
l/L, where L is an expected dwell time between a channel change;
in response to the discounted first confidence score Ch−
l/L being greater than the second confidence score C0, employing, by the system, the first reference descriptor associated with the first match Mh for selecting related content; and
in response to the discounted first confidence score Ch−
l/L not being greater than the second confidence score C0, employing, by the system, the second reference descriptor associated with the second match M0 for selecting the related content.
2 Assignments
0 Petitions
Accused Products
Abstract
Systems, methods, devices, and computer program products provide social and interactive applications for detecting repeating content in broadcast media. In some implementations, a method includes: generating a database of audio statistics from content; generating a query from the database of audio statistics; running the query against the database of audio statistics to determine a non-identity match; if a non-identity match exists, identifying the content corresponding to the matched query as repeating content.
74 Citations
15 Claims
-
1. A method, comprising:
-
determining, by a system including a processor, a first match Mh between a first audio descriptor representing a first recording at a first time step in an environment and a first reference descriptor, the first match Mh having-a first confidence score Ch indicative of a confidence of the first match, the first time step having a time step length l; determining, by the system, a second match M0 between a second audio descriptor representing a second recording at a second time step in the environment and a second reference descriptor, the second match M0 having a second confidence score C0 indicative of a confidence of the second match, the second time step having the time step length l, wherein the first time step is prior to the second time step, and the first match Mh and the second match M0 are non-identity matches determined using a direct or locality sensitive hashing function and a validation process to select a most accurate match out of a plurality of candidate matches, wherein the first confidence score Ch and the second confidence score C0 are based upon a log-likelihood function given by an audio fingerprinting process; discounting, by the system, the first confidence score Ch by a discount value l/L to generate a discounted first confidence score Ch−
l/L, where L is an expected dwell time between a channel change;in response to the discounted first confidence score Ch−
l/L being greater than the second confidence score C0, employing, by the system, the first reference descriptor associated with the first match Mh for selecting related content; andin response to the discounted first confidence score Ch−
l/L not being greater than the second confidence score C0, employing, by the system, the second reference descriptor associated with the second match M0 for selecting the related content. - View Dependent Claims (2, 3, 4, 5)
-
-
6. A non-transitory computer-readable medium having instructions stored thereon that, in response to execution, cause a system including a processor to perform operations comprising:
-
determining a first match Mh between a first audio descriptor representing a first recording at a first time step in an environment and a first reference descriptor, the first match Mh having-a first confidence score Ch indicative of a confidence of the first match, the first time step having a time step length l; determining a second match M0 between a second audio descriptor representing a second recording at a second time step in the environment and a second reference descriptor, the second match M0 having a second confidence score C0 indicative of a confidence of the second match, the second time step having the time step length l, wherein the first time step is temporally prior to the second time step, and the first match Mh and the second match M0 are non-identity matches determined using a direct or locality sensitive hashing function and a validation process to select a most accurate match out of a plurality of candidate matches, wherein the first confidence score Ch and the second confidence score C0 are based upon a log-likelihood function given by an audio fingerprinting process; discounting the first confidence score Ch by a discount value l/L to generate a discounted first confidence score Ch−
l/L, where L is an expected dwell time between a channel change;in response to the discounted first confidence score Ch−
l/L being greater than the second confidence score C0, employing the first reference descriptor associated with the first match Mh for selecting related content; andin response to the discounted first confidence score Ch−
l/L not being greater than the second confidence score C0, employing the second reference descriptor associated with the second match M0 for selecting the related content. - View Dependent Claims (7, 8, 9, 10)
-
-
11. A system, comprising:
-
a processor, communicatively coupled to a memory that stores computer-executable instructions, that executes or facilitates execution of the computer-executable instructions to perform operations comprising; a social application server; determine a first match Mh between a first audio descriptor representing a first recording at a first time step in an environment and a first reference descriptor, the first match Mh having-a first confidence score Ch indicative of a confidence of the first match, the first time step having a time step length l; determine a second match M0 between a second audio descriptor representing a second recording at a second time step in the environment and a second reference descriptor, the second match M0 having a second confidence score C0 indicative of a confidence of the second match, the second time step having the time step length l, wherein the first time step is temporally prior to the second time step, and the first match Mh and the second match M0 are non-identity matches determined using a direct or locality sensitive hashing function and a validation process to select a most accurate match out of a plurality of candidate matches, wherein the first confidence score Ch and the second confidence score C0 are based upon a log-likelihood function given by an audio fingerprinting process; discount the first confidence score Ch by a discount value l/L to generate a discounted first confidence score Ch−
l/L, where L is an expected dwell time between a channel change;in response to the discounted first confidence score Ch−
l/L being greater than the second confidence score C0, employ the first reference descriptor associated with the first match Mh for selecting related content; andin response to the discounted first confidence score Ch−
l/L not being greater than the second confidence score C0, employ the second reference descriptor associated with the second match M0 for selecting the related content. - View Dependent Claims (12, 13, 14, 15)
-
Specification