METHOD AND SYSTEM FOR EXTRACTING VIDEO/AUDIO FRAMES ASSOCIATED WITH CHARACTERS OR SUB-PLOTS FROM A MEDIA FILE
Publication Date: 2012-Sep-17
The IP.com Prior Art Database
A method and system is disclosed for extracting one or more video/audio frames associated with one or more characters or sub-plots from a media file. Thereafter, the one or more video/audio frames are combined to form a new media file for a user.
Page 01 of 2
METHOD AND SYSTEM FOR EXTRACTING VIDEO /AUDIO FRAMES ASSOCIATED WITH CHARACTERS OR SUB-PLOTS FROM A MEDIA FILE
Disclosed is a method and system for extracting one or more video/audio frames associated with one or more characters or sub-plots from a media file. Thereafter, the one or more video/audio frames are combined to form a new media file for a user.
In accordance with the method and system disclosed herein, a user is provided with an option to select one or more of a scene and a character from a story. In an exemplary implementation, a user is enabled to click a frame on a media file of a live broadcast. An interface such as a button is provided for the user to press or click while selecting a character or scene in the frame of the media file of the story. Based on the user selection, the frame containing is assigned an identifier and stored in a database.
Subsequently, a recording system searches for one or more related frames. In an exemplary implementation, facial recognition and video recognition techniques are used to find one or more related frames in the media file/ live-stream. Similarly, voice recognition techniques are used to identify one or more characters. Further, to determine story/sub-plot, audio is converted to text and the text is analyzed to determine whether a conversation between the character (selected by the user) and another character is part of the same story. A story relationship would be presumed if any of the following were true of other scenes in the selected media stream:
Facial recognition technique discovered the character found in the frame
selected by the user was also present in another frame
Facial recognition technique discovered a person in another frame who was also
found in the selected frame but was not the selected character
Facial recognition technique discovered other person(s) in another frame who
were also found in step 2 but are not the selected character
The background scenery or location i...