Identifying stories in media content

Patent metadata

Field	Value
Publication number	US-9734408-B2
Application number	US-201314905487-A
Country	US
Kind code	B2
Filing date	Jul 18, 2013
Priority date	Jul 18, 2013
Publication date	Aug 15, 2017
Grant date	Aug 15, 2017

Abstract

Techniques associated with identifying stories in media content are described in various implementations. In one example implementation, a method may include receiving media content, and segmenting the media content into a plurality of media segments based on auditory indicators included in the media content. The method may also include analyzing a first media segment to determine a first set of concepts associated with the first media segment, and analyzing a second media segment to determine a second set of concepts associated with the second media segment. The method may further include comparing the first set of concepts to the second set of concepts to determine a conceptual similarity between the first set of concepts and the second set of concepts, and in response to the conceptual similarity exceeding a similarity threshold, identifying a story that includes the first media segment and the second media segment.

First claim

What is claimed is:

1. A computer-implemented method of identifying stories in media content, the method comprising:
receiving, at a computing system, media content;
segmenting, using the computing system, the media content into a plurality of media segments based on auditory indicators included in the media content;
analyzing, using the computing system, a first media segment from among the plurality of media segments to determine a first set of concepts associated with the first media segment;
analyzing, using the computing system, a second media segment from among the plurality of media segments to determine a second set of concepts associated with the second media segment;
comparing, using the computing system, the first set of concepts to the second set of concepts to determine a first conceptual similarity between the first set of concepts and the second set of concepts; and
in response to the first conceptual similarity exceeding a similarity threshold, identifying a story that includes the first media segment and the second media segment, wherein identifying comprises:
analyzing an intervening media segment between the first media segment and the second media segment to determine a third set of concepts associated with the intervening media segment;
determining that both a second conceptual similarity between the third set of concepts associated with the intervening media segment and the first set of concepts and a third conceptual similarity between the third set of concepts and the second set of concepts do not exceed the similarity threshold; and
merging the first media segment, the second media segment, and the intervening media segment even though the second and third conceptual similarities do not exceed the similarity threshold.

2. The computer-implemented method of claim 1 , wherein segmenting the media content is further based on visual indicators included in the media content.

3. The computer-implemented method of claim 1 , wherein the conceptual similarity is determined based on matching concepts included in the first set to concepts included in the second set.

4. The computer-implemented method of claim 1 , wherein the conceptual similarity is determined based on conceptual distances between concepts included in the first set and concepts included in the second set.

5. The computer-implemented method of claim 1 , wherein the media content comprises a multimedia content stream.

6. The method of claim 1 , further comprising identifying the story that includes the first media segment, the second media segment, and the intervening media segment as a non-story responsive to determining the duration of the story is less than a minimum time threshold.

7. The method of claim 1 , further comprising identifying the story that includes the first media segment, the second media segment, and the intervening media segment as a non-story responsive to determining the story includes advertisement or non-story segue concepts.

8. A story recognition system comprising:
one or more processors; and
a non-transitory computer-readable medium storing instructions that, when executed by the one or more processors, cause the story recognition system to:
segment media content into a plurality of media segments based on auditory indicators included in the media content;
analyze the media segments to determine concepts associated with the respective media segments;
compare the concepts associated with a first media segment to the concepts associated with a second media segment to determine a first conceptual similarity between the first media segment and the second media segment;
obtain a story that includes the first media segment and the second media segment in response to a determination that the first conceptual similarity exceeds a similarity threshold, wherein obtaining the story comprises:
analyzing an intervening media segment between the first media segment and the second media segment to determine a third set of concepts associated with the intervening media segment;
determining that both a second conceptual similarity between the third set of concepts associated with the intervening media segment and the first set of concepts and a third conceptual similarity between the third set of concepts and the second set of concepts do not exceed the similarity threshold; and
merging the first media segment, the second media segment, and the intervening media segment into the story even though the second and third conceptual similarities do not exceed the similarity threshold.

9. The story recognition system of claim 8 , wherein the instructions, when executed by the one or more processors, cause the story recognition engine to segment the media content further based on visual indicators included in the media content.

10. The story recognition system of claim 8 , wherein the conceptual similarity is determined based on matching the concepts associated with the first media segment to the concepts associated with the second media segment.

11. The story recognition system of claim 8 , wherein the conceptual similarity is determined based on conceptual distances between the concepts associated with the first media segment and the concepts associated with the second media segment.

12. The story recognition system of claim 8 , wherein the media content comprises a multimedia content stream.

13. The story recognition system of claim 8 , wherein the non-transitory machine-readable medium further stores instructions that, when executed by the one or more processors, cause the story recognition system to:
identify the story that includes the first media segment, the second media segment, and the intervening media segment as a non-story responsive to determining the duration of the story is less than a minimum time threshold.

14. The story recognition system of claim 8 , wherein the non-transitory machine-readable medium further stores instructions that, when executed by the one or more processors, cause the story recognition system to:
identify the story that includes the first media segment, the second media segment, and the intervening media segment as a non-story responsive to determining the story includes advertisement or non-story segue concepts.

15. A non-transitory computer-readable storage medium storing instructions that, when executed by one or more processors, cause the one or more processors to:
segment media content into a plurality of media segments based on auditory indicators included in the media content;
analyze the media segments to determine concepts associated with the respective media segments;
compare the concepts associated with a first media segment to the concepts associated with a second media segment to determine a first conceptual similarity between the first media segment and the second media segment;
obtain a story that includes the first media segment and the second media segment in response to a determination that the first conceptual similarity exceeds a similarity threshold, wherein obtaining the story comprises:
analyzing an intervening media segment between the first media segment and the second media segment to determine a third set of concepts associated with the intervening media segment;
determining that both a second conceptual similarity between the third set of concepts associated with the intervening media segment and the first set of concepts and a third conceptual similarity between the third set of concepts and the second set of concepts do not exceed the similarity threshold; and
merging the first media segment, the second media segment, and the intervening m

Assignees

• Longsand Ltd

Inventors

• Betley Abigail
• Ayo Aresti Unai
• Toone David

Classifications

• H04N21/8541

  involving branching, e.g. to different story endings · CPC title

• H04N21/23424

  involving splicing one content stream with another content stream, e.g. for inserting or substituting an advertisement · CPC title

• G11B27/10

  Indexing; Addressing; Timing or synchronising; Measuring tape travel · CPC title

• G06K2009/00738

  Physics · mapped topic

• G06K9/00751 Primary

  Physics · mapped topic