Video Summarization using Speech Recognition and Text Summarization
A two-fold approach to fetch the subject matter of videos through effective summarization using an Automatic Speech Recognition system based on a Convolutional Neural Network and Extractive Text Summarization.
Abstract
Videos on the internet have been increasingly becoming the chief source of knowledge and information in today’s digital age. However, with increasing length of videos and diminishing time to spare in everyone’s lives, a need has emerged for Video Summarization tools that can provide a good summary about the content of videos without the need to watch videos in their entirety. In this paper, we introduce a two-fold approach to fetch the subject matter of videos through effective summarization. The employed approach comprises of two phases: the first phase involves performing speech-to-text conversion using an Automatic Speech Recognition(ASR) system based on a Convolutional Neural Network(CNN) for generating respective transcripts for input videos while the second phase involves performing Extractive Text Summarization to summarize the text generated by extracting the important information.