Abstract: The present invention provides a method and system for automatically reframing and transforming videos to different aspect ratios, such that the source video has a fixed resolution and multiple outputs of different aspect ratios are generated. The present invention automatically reframes and transforms horizontal videos into vertical, portrait, square, and/or landscape for social media distribution. The method comprises acquiring input video details; extracting audio; detecting shot change points and extracting frames; detecting salient regions; detecting text; generating and stabilizing viewports; applying genre-specific transformations and generating transformed images; recreating text; obtaining output frames; generating videos and uploading the reframed and transformed videos into cloud. The present invention helps to reframe and transform videos by maintaining the visibility of regions of interest.

Abstract: A system and a method for automatically identifying key dialogues in media is disclosed herein. In the method disclosed herein, the key dialogues engine receives the media asset and extract transcript data and supplementary data. The key dialogues engine processes the transcript data into a plurality of transcript data elements and associate the transcript data elements with respective data elements selected from the supplementary data.

Abstract: A system and a method for automatically identifying key dialogues in media is disclosed herein. In the method disclosed herein, the key dialogues engine receives the media asset and extract transcript data and supplementary data. The key dialogues engine processes the transcript data into a plurality of transcript data elements and associate the transcript data elements with respective data elements selected from the supplementary data.