Sound Approach for AI Generated Video Highlights

A large number of cameras record video around the clock, producing huge quantities of data. Processing or scrutinizing such huge chunks of videos and extracting salient events from the large video databases manually is a quite laborious and time-consuming task. Again such video data requires huge memory space to store. Due to such an increase in video data, there is a need for effective techniques for video retrieval and analysis as no one has enough time to watch the entire video. Generating highlights from the video is called as video summarization and this technique produce a short, informative summary of a long video without losing the important data frames. This technique is useful in the indexing and classification of such videos in the video database. This paper provides a review of different approaches given by various researchers in the field of video summarization.