Method and apparatus for analyzing media content

Patent metadata

Field	Value
Publication number	US-9898685-B2
Application number	US-201414264183-A
Country	US
Kind code	B2
Filing date	Apr 29, 2014
Priority date	Apr 29, 2014
Publication date	Feb 20, 2018
Grant date	Feb 20, 2018

Abstract

Aspects of the subject disclosure may include, for example, a method for determining a first set of features in first images of first media content, generating a similarity score by processing the first set of features with a favorability model derived by identifying generative features and discriminative features of second media content that is favored by a viewer, and providing the similarity score to a network for predicting a response by the viewer to the first media content. Other embodiments are disclosed.

First claim

What is claimed is:

1. A device, comprising:
a processing system including a processor; and
a memory that stores executable instructions that, when executed by the processing system, facilitate performance of operations, comprising:
scanning a first plurality of images of first audio visual media content to create a first raw catalog of image objects extracted from the first plurality of images;
filtering the first raw catalog of image objects using first filtering criteria to select a first plurality of relevant image objects, wherein the first filtering criteria comprise a number of occurrences of an image object, a running time of occurrence of the image object, or a combination thereof;
scanning and filtering a second plurality of images of a second audio visual media content using second filtering criteria to identify a second plurality of image objects, wherein the second audio visual media content has been previously experienced by a viewer, wherein the second audio visual media content is selected from a set of content being viewed by the viewer within a time period, wherein the second audio visual media content is of a same genre as the first audio visual media content, and wherein one or both of the first filtering criteria and the second filtering criteria include information from a viewer profile;
comparing the first plurality of relevant image objects to the second plurality of image objects to identify generative visual features and discriminative visual features with respect to the first audio visual media content and the second audio visual media content;
determining a set of similarity matrices according to the generative visual features and the discriminative visual features;
processing the similarity matrices to generate a similarity score according to correlated visual features of a set of previously-experienced audio visual media content; and
generating a recommendation for the first audio visual media content according to the similarity score.

2. The device of claim 1 , wherein the recommendation is further generated according to a favorability rating of the second audio visual media content by the viewer, and wherein the processing system comprises a plurality of processors operating in a distributed processing environment.

3. The device of claim 1 , wherein the similarity matrices are determined based on media content other than textual content.

4. The device of claim 1 , wherein the operations further comprise:
extracting two-dimensional visual features from the first plurality of relevant image objects; and
extracting timing information associated with the two-dimensional visual features to generate three-dimensional visual features.

5. The device of claim 1 , wherein the operations further comprise comparing the first plurality of relevant image objects to a third plurality of image objects in a third plurality of images of third audio visual media content to identify second generative visual features and second discriminative visual features with respect to the first audio visual media content and the third audio visual media content, wherein the third audio visual media content has been previously experienced by the viewer, and wherein the second audio visual media content and the third audio visual media content comprise videos from a category, time period, genre, or combinations thereof.

6. The device of claim 1 , wherein the operations further comprise:
scanning a plurality of training images of a plurality of training media content associated with the viewer to detect a plurality of visual training image objects present within the plurality of training images;
comparing training images of the plurality of training images to identify a plurality of training generative visual features and a plurality of training discriminative visual features with respect to combinations of the training media content; and
generating a model of correlated image objects according to the plurality of training generative visual features and the plurality of training discriminative visual features.

7. The device of claim 6 , wherein the operations further comprise accessing a plurality of favorability ratings for the training media content, wherein the model of correlated image objects is further generated according to the plurality of favorability ratings.

8. The device of claim 1 , wherein the operations further comprise receiving an indication of a viewing decision responsive to the recommendation.

9. The device of claim 8 , wherein the operations further comprise transmitting the first audio visual media content according to the indication.

10. The device of claim 9 , wherein the operations further comprise:
identifying a first plurality of objects in the first plurality of relevant image objects by comparing the first plurality of relevant image objects to a database of known objects; and
comparing the first plurality of objects to a second plurality of objects of the second audio visual media content to identify generative objects and discriminative objects with respect to the first media content and the second media content, wherein the similarity matrices are further determined according to the generative objects and the discriminative objects.

11. The device of claim 10 , wherein one of the first plurality of objects comprises a face, a person, or any combinations thereof.

12. The device of claim 10 , wherein the first plurality of objects comprise a person, that is identified according to a facial recognition algorithm.

13. A non-transitory machine-readable storage medium, comprising executable instructions that, when executed by a processing system including a processor, facilitate performance of operations, comprising:
filtering a first raw catalog of image objects derived from scanning of first audio visual media content to select a first plurality of relevant image objects from the first raw catalog of image objects, the filtering being performed using first criteria, wherein the first criteria comprise a number of occurrences of an image object, a running time of occurrence of the image object, or a combination thereof;
identifying second relevant image objects derived from scanning a plurality of audio visual media content items, the identifying being performed using second criteria, wherein the plurality of audio visual media content items have been experienced by a viewer, wherein the plurality of audio visual media content are selected from a set of content being viewed by the viewer within a time period, wherein the plurality of audio visual media content are of a same genre as the first audio visual media content, and wherein one or both of the first criteria and the second criteria include information from a viewer profile;
comparing the first plurality of relevant image objects of the first audio visual media content to the second relevant image objects to identify common visual features and distinct visual features with respect to the first audio visual media content and the plurality of audio visual media content items;
generating a similarity score according to the common visual features and the distinct visual features; and
transmitting the first audio visual media content to a device if the similarity score exceeds a threshold.

14. The non-transitory machine-readable storage medium of claim 13 , wherein the operations further comprise generating a recommendation for the first audio visual media content according to the similarity score, and wherein the processing system comprises a plurality of processors operating in a distributed processing environment.

15. The n

Assignees

• At & T Ip I Lp
• At & T Ip I Lp

Inventors

• Gopalan Raghuraman

Classifications

• H04N21/251 Primary

  Learning process for intelligent management, e.g. learning user preferences for recommending movies (details of learning user preferences for the retrieval of video data in a video database G06F16/739; computer systems using learning methods G06N3/08) · CPC title

• G06F18/22

  Matching criteria, e.g. proximity measures · CPC title

• G06F18/2413

  based on distances to training or reference patterns · CPC title

• G06F17/30784

  Physics · mapped topic

• G06K9/00624

  Physics · mapped topic