Iterative, maximally probable, batch-mode commercial detection for audiovisual content
First Claim
1. A method for detecting one or more commercial breaks in a set of audiovisual content spanning a duration of time, each commercial break including one or more commercials, the method comprising the steps of:
- identifying candidate times within the duration of time spanned by the set of audiovisual content based on an evaluation of one or more cues identified in the audiovisual content, each candidate time representing a possible starting and/or ending time of a commercial;
assigning a score to each candidate time;
evaluating, for each of one or more candidate times,
1) one or more secondary cues that are each different from the one or more cues used to identify the candidate time, and/or
2) the relationship between the candidate time and one or more other candidate times, wherein the score assigned to the candidate time can be adjusted based on the evaluation; and
constructing the one or more commercial breaks based on an evaluation of
1) the scores of the candidate times after the step of evaluating and
2) a relationship between, or relationships among, the candidate times, including by;
comparing a relationship between, or relationships among, the candidate times to one or more probability models that specify an expected relationship between, or relationships among, commercial start and/or end times;
selecting the candidate time with the highest adjusted score to be a commercial start and/or end time in a current commercial break;
successively evaluating each candidate time not yet part of a commercial break, in order of decreasing adjusted score, for possible inclusion in the current commercial break as a commercial start and/or end time, wherein the evaluation of each candidate time for possible inclusion in the current commercial break comprises the steps of;
determining whether each probability of the temporal separation between the candidate time being evaluated and a candidate time already included in the current commercial break is above a specified threshold value;
determining whether the probability of the duration of the current commercial break, if the candidate time being evaluated is added to the current commercial break, is above a specified threshold value; and
determining whether each probability of the temporal separation between the candidate time being evaluated and an already existing commercial break, if any, is above a specified threshold value, wherein if each of the three probabilities is above the corresponding specified threshold value, the candidate time being evaluated is added to the current commercial break;
determining whether there are candidate times having an adjusted score above a specified threshold value that have not yet been included in a commercial break and have not yet been excluded from inclusion in a commercial break;
if there are candidate times having an adjusted score above the specified threshold value that have not yet been included in a commercial break and have not yet been excluded from inclusion in a commercial break, performing the steps of;
selecting the candidate time having the highest adjusted score above the specified threshold value that has not yet been included in a commercial break and has not yet been excluded from inclusion in a commercial break;
determining whether each probability of the temporal separation between the selected candidate time and an already existing commercial break is above a specified threshold value;
if each probability is above the specified threshold value, performing the steps of;
identifying the selected candidate time as a commercial start and/or end time in a current commercial break; and
repeating the step of successively evaluating; and
if each probability is not above the specified threshold value, performing the steps of;
excluding the selected candidate time from inclusion in a commercial break; and
repeating the step of determining whether there are candidate times having an adjusted score above a specified threshold value that have not yet been included in a commercial break and have not yet been excluded from inclusion in a commercial break; and
if there are no candidate times having an adjusted score above the specified threshold value that have not yet been included in a commercial break and have not yet been excluded from inclusion in a commercial break, performing the step of identifying the start and end time of each commercial break and the start time of each commercial in each commercial break.
3 Assignments
0 Petitions
Accused Products
Abstract
Identification of starting and ending times of commercial breaks and commercials within those commercial breaks found in audiovisual content is disclosed. A solution to a “batch optimization” problem is used in which commercial locations within a set of audiovisual content are detected as a group by choosing a set of commercial locations which optimizes a cost function which can include considerations of, for example, 1) one or more cues, 2) relative locations of commercials within the audiovisual content, and/or 3) probability models based on statistics obtained regarding characteristics of typical commercial and commercial breaks. Optimization can be done over the total set of commercial location decisions, rather than on a per-commercial basis. Additionally, the cost function can be iteratively evaluated and many more types of cues and combinations of cues can be used in detection of commercials.
-
Citations
40 Claims
-
1. A method for detecting one or more commercial breaks in a set of audiovisual content spanning a duration of time, each commercial break including one or more commercials, the method comprising the steps of:
-
identifying candidate times within the duration of time spanned by the set of audiovisual content based on an evaluation of one or more cues identified in the audiovisual content, each candidate time representing a possible starting and/or ending time of a commercial; assigning a score to each candidate time; evaluating, for each of one or more candidate times,
1) one or more secondary cues that are each different from the one or more cues used to identify the candidate time, and/or
2) the relationship between the candidate time and one or more other candidate times, wherein the score assigned to the candidate time can be adjusted based on the evaluation; andconstructing the one or more commercial breaks based on an evaluation of
1) the scores of the candidate times after the step of evaluating and
2) a relationship between, or relationships among, the candidate times, including by;comparing a relationship between, or relationships among, the candidate times to one or more probability models that specify an expected relationship between, or relationships among, commercial start and/or end times; selecting the candidate time with the highest adjusted score to be a commercial start and/or end time in a current commercial break; successively evaluating each candidate time not yet part of a commercial break, in order of decreasing adjusted score, for possible inclusion in the current commercial break as a commercial start and/or end time, wherein the evaluation of each candidate time for possible inclusion in the current commercial break comprises the steps of; determining whether each probability of the temporal separation between the candidate time being evaluated and a candidate time already included in the current commercial break is above a specified threshold value; determining whether the probability of the duration of the current commercial break, if the candidate time being evaluated is added to the current commercial break, is above a specified threshold value; and determining whether each probability of the temporal separation between the candidate time being evaluated and an already existing commercial break, if any, is above a specified threshold value, wherein if each of the three probabilities is above the corresponding specified threshold value, the candidate time being evaluated is added to the current commercial break; determining whether there are candidate times having an adjusted score above a specified threshold value that have not yet been included in a commercial break and have not yet been excluded from inclusion in a commercial break; if there are candidate times having an adjusted score above the specified threshold value that have not yet been included in a commercial break and have not yet been excluded from inclusion in a commercial break, performing the steps of; selecting the candidate time having the highest adjusted score above the specified threshold value that has not yet been included in a commercial break and has not yet been excluded from inclusion in a commercial break; determining whether each probability of the temporal separation between the selected candidate time and an already existing commercial break is above a specified threshold value; if each probability is above the specified threshold value, performing the steps of; identifying the selected candidate time as a commercial start and/or end time in a current commercial break; and repeating the step of successively evaluating; and if each probability is not above the specified threshold value, performing the steps of; excluding the selected candidate time from inclusion in a commercial break; and repeating the step of determining whether there are candidate times having an adjusted score above a specified threshold value that have not yet been included in a commercial break and have not yet been excluded from inclusion in a commercial break; and if there are no candidate times having an adjusted score above the specified threshold value that have not yet been included in a commercial break and have not yet been excluded from inclusion in a commercial break, performing the step of identifying the start and end time of each commercial break and the start time of each commercial in each commercial break. - View Dependent Claims (2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38)
-
-
39. A system for detecting one or more commercial breaks in a set of audiovisual content spanning a duration of time, each commercial break including one or more commercials, the system comprising:
-
means for identifying candidate times within the duration of time spanned by the set of audiovisual content based on an evaluation of one or more cues identified in the audiovisual content, each candidate time representing a possible starting and/or ending time of a commercial; means for assigning a score to each candidate time; means for evaluating, for each of one or more candidate times,
1) one or more secondary cues that are each different from the one or more cues used to identify the candidate time, and/or
2) the relationship between the candidate time and one or more other candidate times, wherein the score assigned to the candidate time can be adjusted based on the evaluation; andmeans for constructing the one or more commercial breaks based on an evaluation of
1) the scores of the candidate times after the step of evaluating and
2) a relationship between, or relationships among, the candidate times, comprising;comparing a relationship between, or relationships among, the candidate times to one or more probability models that specify an expected relationship between, or relationships among, commercial start and/or end times; selecting the candidate time with the highest adjusted score to be a commercial start and/or end time in a current commercial break; successively evaluating each candidate time not yet part of a commercial break in order of decreasing adjusted score, for possible inclusion in the current commercial break as a commercial start and/or end time, wherein the evaluation of each candidate time for possible inclusion in the current commercial break comprises the steps of; determining whether each probability of the temporal separation between the candidate time being evaluated and a candidate time already included in the current commercial break is above a specified threshold value; determining whether the probability of the duration of the current commercial break, if the candidate time being evaluated is added to the current commercial break, is above a specified threshold value; and determining whether each probability of the temporal separation between the candidate time being evaluated and an already existing commercial break, if any, is above a specified threshold value, wherein if each of the three probabilities is above the corresponding specified threshold value, the candidate time being evaluated is added to the current commercial break; determining whether there are candidate times having an adjusted score above a specified threshold value that have not yet been included in a commercial break and have not yet been excluded from inclusion in a commercial break; if there are candidate times having an adjusted score above the specified threshold value that have not yet been included in a commercial break and have not yet been excluded from inclusion in a commercial break, performing the steps of; selecting the candidate time having the highest adjusted score above the specified threshold value that has not yet been included in a commercial break and has not yet been excluded from inclusion in a commercial break; determining whether each probability of the temporal separation between the selected candidate time and an already existing commercial break is above a specified threshold value; if each probability is above the specified threshold value, performing the steps of; identifying the selected candidate time as a commercial start and/or end time in a current commercial break; and repeating the step of successively evaluating; and if each probability is not above the specified threshold value, performing the steps of; excluding the selected candidate time from inclusion in a commercial break; and repeating the step of determining whether there are candidate times having an adjusted score above a specified threshold value that have not yet been included in a commercial break and have not yet been excluded from inclusion in a commercial break; and if there are no candidate times having an adjusted score above the specified threshold value that have not yet been included in a commercial break and have not yet been excluded from inclusion in a commercial break, performing the step of identifying the start and end time of each commercial break and the start time of each commercial in each commercial break.
-
-
40. A computer readable storage medium or media encoded with one or more computer programs including instructions for detecting one or more commercial breaks in a set of audiovisual content spanning a duration of time, each commercial break including one or more commercials, the one or more computer programs comprising:
-
instructions for identifying candidate times within the duration of time spanned by the set of audiovisual content based on an evaluation of one or more cues identified in the audiovisual content, each candidate time representing a possible starting and/or ending time of a commercial; instructions for assigning a score to each candidate time; instructions for evaluating, for each of one or more candidate times,
1) one or more secondary cues that are each different from the one or more cues used to identify the candidate time, and/or
2) the relationship between the candidate time and one or more other candidate times, wherein the score assigned to the candidate time can be adjusted based on the evaluation; andinstructions for constructing the one or more commercial breaks based on an evaluation of
1) the scores of the candidate times after the step of evaluating and
2) a relationship between, or relationships among, the candidate times, comprising;comparing a relationship between, or relationships among, the candidate times to one or more probability models that specify an expected relationship between, or relationships among, commercial start and/or end times; selecting the candidate time with the highest adjusted score to be a commercial start and/or end time in a current commercial break; successively evaluating each candidate time not yet part of a commercial break, in order of decreasing adjusted score, for possible inclusion in the current commercial break as a commercial start and/or end time, wherein the evaluation of each candidate time for possible inclusion in the current commercial break comprises the steps of; determining whether each probability of the temporal separation between the candidate time being evaluated and a candidate time already included in the current commercial break is above a specified threshold value; determining whether the probability of the duration of the current commercial break, if the candidate time being evaluated is added to the current commercial break, is above a specified threshold value; and determining whether each probability of the temporal separation between the candidate time being evaluated and an already existing commercial break, if any, is above a specified threshold value, wherein if each of the three probabilities is above the corresponding specified threshold value, the candidate time being evaluated is added to the current commercial break; determining whether there are candidate times having an adjusted score above a specified threshold value that have not yet been included in a commercial break and have not yet been excluded from inclusion in a commercial break; if there are candidate times having an adjusted score above the specified threshold value that have not yet been included in a commercial break and have not yet been excluded from inclusion in a commercial break, performing the steps of; selecting the candidate time having the highest adjusted score above the specified threshold value that has not yet been included in a commercial break and has not yet been excluded from inclusion in a commercial break; determining whether each probability of the temporal separation between the selected candidate time and an already existing commercial break is above a specified threshold value; if each probability is above the specified threshold value, performing the steps of; identifying the selected candidate time as a commercial start and/or end time in a current commercial break; and repeating the step of successively evaluating; and if each probability is not above the specified threshold value, performing the steps of; excluding the selected candidate time from inclusion in a commercial break; and repeating the step of determining whether there are candidate times having an adjusted score above a specified threshold value that have not yet been included in a commercial break and have not yet been excluded from inclusion in a commercial break; and if there are no candidate times having an adjusted score above the specified threshold value that have not yet been included in a commercial break and have not yet been excluded from inclusion in a commercial break, performing the step of identifying the start and end time of each commercial break and the start time of each commercial in each commercial break.
-
Specification