Developing a Video Summarizing Tool using Machine Learning

Abstract

In recent years, video has become a highly significant form of visual data, and the explosion of short video platforms like TikTok, Instagram, and Facebook has led people to prefer to consume short content than watching a long video without skipping any second. To solve this problem, a video summarizing system was developed in this work, to compress a long video into a much shorter version, but still preserving the important contents of the original one. As a trend in the world nowadays, Artificial Intelligence was used to make the process more efficient, and it proved to be perfectly suitable for handling this task too. The primary goal of the project was to do research in the field of video summarization and develop a tool that can assist people in improving video summarization workflow by significantly reducing processing time and simplifying video summarizing process. The original research focused on improving the performance of the method but not on the usability for normal users. With the application developed in this project, the user can summarize a video just by inputting the original video file and getting a summarized version of it in terms of seconds.