An Approach for Automatic Segmentation of Scenes in Educational Videos through the use of Audio Transcription and Semantic Annotation

Abstract

In recent years, educational videos are becoming more and more popular. Due to this increase in the amount of didactic content in the video format present on the web, it is interesting to make it possible for a search term to be related to a specific segment of the video. Better navigability allows the user to have quicker access to the topics that interest him, avoiding irrelevant content. This article proposes a method for automatic segmentation of scenes in educational videos through the use of automatic audio transcription and semantic annotation. With this targeting, you can improve content search on these videos by improving the user experience on e-learning platforms or educational video repositories.