Caption and/or Metadata Synchronization for Replay of Previously or Simultaneously Recorded Live Programs
A synchronization process between captioning data and/or corresponding metatags and the associated media file parses the media file, correlates the caption information and/or metatags with segments of the media file, and provides a capability for textual search and selection of particular segments. A time-synchronized version of the captions is created that is synchronized to the moment that the speech is uttered in the recorded media. The caption data is leveraged to enable search engines to index not merely the title of a video, but the entirety of what was said during the video as well as any associated metatags relating to contents of the video. Further, because the entire media file is indexed, a search can request a particular scene or occurrence within the event recorded by the media file, and the exact moment within the media relevant to the search can be accessed and played for the requester.
1 . A method in a computer system for caption synchronization of media programs, comprising:
receiving, utilizing at least one processing unit, at least a portion of a media stream and at least a portion of a caption data stream, the media stream and the caption data stream corresponding to an event, the media stream comprising at least an audio component;
tuning, utilizing the least one processing unit, an audio mining engine with the at least the portion of the caption data stream;
producing, utilizing the at least one processing unit, an audio mining transcript including timing information by executing the audio mining engine on the at least the portion of the media stream;
aligning, utilizing the at least one processing unit, the audio mining transcript with the at least the portion of the caption data stream; and
generating a time synchronized caption data stream, utilizing the at least one processing unit, by applying the timing information from the audio mining transcript to the at least the portion of the caption stream based on the aligning.