Claims:We claim:
1. A method of generating a contextual summary of multimedia content, the method comprising:
retrieving, by a contextual summary generator (107), the multimedia content comprising one or more scenes from a multimedia content database (105);
generating, by the contextual summary generator (107), one or more scene descriptors, describing a scene, for each of the one or more scenes;
identifying, by the contextual summary generator (107), an emotion factor in each of the one or more scenes based on each of the one or more scene descriptors, each of one or more speech descriptors and each of one or more textual descriptors associated with each of the one or more scenes;
generating, by the contextual summary generator (107), a context descriptor, indicating context of each of the one or more scenes, for each of the one or more scenes based on analysis of each emotion factor in each of the one or more scenes and non-speech descriptors associated with each of the one or more scenes; and
correlating, by the contextual summary generator (107), the one or more scene descriptors, the one or more textual descriptors, and the context descriptors of the one or more scenes based on a dynamically configured threshold value to generate the contextual summary of the multimedia content.
2. The method as claimed in claim 1, wherein the one or more scene descriptors are generated by:
determining one or more key frames from each of the one or more scenes using one or more predefined techniques;
obtaining one or more non-redundant key frames from the one or more key frames based on distance between each of the one or more key frames; and
identifying in each of the one or more non-redundant key frames, one or more objects, one or more actions related to the objects and background where the one or more objects are present to generate one or more scene descriptors related to each of the one or more scenes.
3. The method as claimed in claim 1 further comprises assigning a rank to each of the one or more scenes based on each of the one or more scene descriptors, the context descriptors and the dynamically generated weightage corresponding to each of the one or more scenes.

4. The method as claimed in claim 3, wherein the dynamically generated weightage is computed based on frequency and duration of occurrence of one or more objects in each of the one or more scenes.

5. The method as claimed in claim 1, wherein the one or more scenes selected for correlating is based on a rank.

6. The method as claimed in claim 1, wherein the one or more speech descriptors comprise information related to at least one of pitch, tone or speed of speech audio signals present in each of the one or more scenes.

7. The method as claimed in claim 1, wherein the one or more textual descriptors are generated by converting speech audio signals in each of the one or more scenes into a text format.

8. The method as claimed in claim 1, wherein the contextual summary is in at least one of an audio format, a video format, an image format or a textual format.

9. The method as claimed in claim 1, wherein the context descriptor for each of the one or more scenes describes at least one of a formal context, an informal context, a humorous context, a serious context, a positive context or a negative context.

10. The method as claimed in claim 1, wherein the dynamically configured threshold is based on at least one of time for which the contextual summary of the multimedia content should be generated and information to be present in the contextual summary of the multimedia content, wherein the information indicates the one or more scene descriptors, the one or more speech descriptors, the one or more non-speech descriptors, the one or more textual descriptors and the one or more context descriptors.

11. A contextual summary generator (107) for generating a contextual summary of multimedia content, the contextual summary generator (107) comprising:
a processor (109); and
a memory (113) communicatively coupled to the processor (109), wherein the memory (113) stores the processor (109)-executable instructions, which, on execution, causes the processor (109) to:
retrieve the multimedia content comprising one or more scenes from a multimedia content database (105);
generate one or more scene descriptors, describing a scene, for each of the one or more scenes;
identify an emotion factor in each of the one or more scenes based on each of the one or more scene descriptors, each of one or more speech descriptors and each of one or more textual descriptors associated with each of the one or more scenes;
generate a context descriptor, indicating context of each of the one or more scenes, for each of the one or more scenes based on analysis of each emotion factor in each of the one or more scenes and non-speech descriptors associated with each of the one or more scenes; and
correlate the one or more scene descriptors, the one or more textual descriptors, and the context descriptors of the one or more scenes based on a dynamically configured threshold value to generate the contextual summary of the multimedia content.
12. The contextual summary generator (107) as claimed in claim 11, wherein to generate the one or more scene descriptors, the instructions cause the processor (109) to:
determine one or more key frames from each of the one or more scenes using one or more predefined techniques;
obtain one or more non-redundant key frames from the one or more key frames based on distance between each of the one or more key frames; and
identify in each of the one or more non-redundant key frames, one or more objects, one or more actions related to the objects and background where the one or more objects are present to generate one or more scene descriptors related to each of the one or more scenes.
13. The contextual summary generator (107) as claimed in claim 11, wherein the processor (109) is further configured to assign a rank to each of the one or more scenes based on each of the one or more scene descriptors, the context descriptors and the dynamically generated weightage corresponding to each of the one or more scenes.

14. The contextual summary generator (107) as claimed in claim 13, wherein the processor (109) computes the dynamically generated weightage based on frequency and duration of occurrence of one or more objects in each of the one or more scenes.

15. The contextual summary generator (107) as claimed in claim 11, wherein the processor (109) selects the one or more scenes for correlating based on a rank.

16. The contextual summary generator (107) as claimed in claim 11, wherein the one or more speech descriptors comprise information related to at least one of pitch, tone or speed of speech audio signals present in each of the one or more scenes.

17. The contextual summary generator (107) as claimed in claim 11, wherein the processor (109) generates the one or more textual descriptors by converting speech audio signals in each of the one or more scenes into a text format.

18. The contextual summary generator (107) as claimed in claim 11, wherein the contextual summary is in at least one of an audio format, a video format, an image format or a textual format.

19. The contextual summary generator (107) as claimed in claim 11, wherein the context descriptor for each of the one or more scenes describes at least one of a formal context, an informal context, a humorous context, a serious context, a positive context or a negative context.

20. The contextual summary generator (107) as claimed in claim 11, wherein the dynamically configured threshold is based on at least one of time for which the contextual summary of the multimedia content should be generated and information to be present in the contextual summary of the multimedia content, wherein the information indicates the one or more scene descriptors, the one or more speech descriptors, the one or more non-speech descriptors, the one or more textual descriptors and the one or more context descriptors.

Dated this 18th day of May 2017

SWETHA S.N
OF K & S PARTNERS
AGENT FOR THE APPLICANT
IN/PA-2123
, Description:TECHNICAL FIELD
The present subject matter relates generally to paraphrasing multimedia content, and more particularly, but not exclusively to a method and a system for generating a contextual summary of multimedia content.