Back to results
US 2012/0059845 A1Published application
Detecting Repeating Content In Broadcast Media
AssigneeGoogle Inc.
InventorsMichele Covell, Shumeet Baluja, Michael Fink
Life of this patent
Estimating the term…
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.
Claims39
A method performed by one or more computers, comprising:
- receiving a reference vector including multiple reference audio descriptors for identifying audio content;
- receiving a query vector including multiple query audio descriptors of an audio snippet;
- identifying a match between the query vector and the reference vector, including determining a match probability between the query vector and the reference vector based at least in part on:
- a difference score of each query audio descriptor indicating a difference between the query audio descriptor and a reference audio descriptor ;
- and
- a conditional probability that when a query audio descriptor indicates that the query audio descriptor is interference, a subsequent query audio descriptor is also interference;
- and
- identifying that the audio snippet is a snippet of the audio content based on the match.
5 more elements
A method executed by a computer, the method comprising:
- determining a first match between an audio descriptor and a first reference descriptor, the first match occurring at a first time step and being associated with a first confidence score, the first time step having a time step length;
- determining a second match between the audio descriptor and a second reference descriptor, the second match occurring at a second time step and being associated with a second confidence score, the second time step having the time step length;
- discounting the second confidence score by a discount value, the discount value being calculated based at least in part on a ratio between the time step length and an expected dwell time;
- determining that the discounted second confidence score is greater than the first confidence score;
- and
- identifying the first reference descriptor and the second reference descriptor as matching based on the determining.
3 more elements
A computer program product stored on a non-transitory storage device, operable to cause one or more computers to perform operations comprising:
- receiving a reference vector including multiple reference audio descriptors for identifying audio content;
- receiving a query vector including multiple query audio descriptors of an audio snippet;
- identifying a match between the query vector and the reference vector, including determining a match probability between the query vector and the reference vector based at least in part on:
- a difference score of each query audio descriptor indicating a difference between the query audio descriptor and a reference audio descriptor;
- and
- a conditional probability that when a query audio descriptor indicates that the query audio descriptor is interference, a subsequent query audio descriptor is also interference;
- and
- identifying that the audio snippet is a snippet of the audio content based on the match.
5 more elements
A computer program product stored on a non-transitory storage device, operable to cause one or more computers to perform operations, the method comprising:
- determining a first match between an audio descriptor and a first reference descriptor, the first match occurring at a first time step and being associated with a first confidence score, the first time step having a time step length;
- determining a second match between the audio descriptor and a second reference descriptor, the second match occurring at a second time step and being associated with a second confidence score, the second time step having the time step length;
- discounting the second confidence score by a discount value, the discount value being calculated based at least in part on a ratio between the time step length and an expected dwell time;
- determining that the discounted second confidence score is greater than the first confidence score;
- and
- identifying the first reference descriptor and the second reference descriptor as matching based on the determining.
3 more elements
A system, comprising:
- one or more computers configured to perform operations comprising:
- receiving a reference vector including multiple reference audio descriptors for identifying audio content;
- receiving a query vector including multiple query audio descriptors of an audio snippet;
- identifying a match between the query vector and the reference vector, including determining a match probability between the query vector and the reference vector based at least in part on:
- a difference score of each query audio descriptor indicating a difference between the query audio descriptor and a reference audio descriptor;
- and
- a conditional probability that when a query audio descriptor indicates that the query audio descriptor is interference, a subsequent query audio descriptor is also interference;
- and
- identifying that the audio snippet is a snippet of the audio content based on the match.
6 more elements
A system, comprising:
- one or more computers configured to perform operations comprising:
- determining a first match between an audio descriptor and a first reference descriptor, the first match occurring at a first time step and being associated with a first confidence score, the first time step having a time step length;
- determining a second match between the audio descriptor and a second reference descriptor, the second match occurring at a second time step and being associated with a second confidence score, the second time step having the time step length;
- discounting the second confidence score by a discount value, the discount value being calculated based at least in part on a ratio between the time step length and an expected dwell time;
- determining that the discounted second confidence score is greater than the first confidence score;
- and
- identifying the first reference descriptor and the second reference descriptor as matching based on the determining.
4 more elements
Dependencies are inferred from the claim text.