Extracting Entities from Video and Tagging them - NLP Model

Abstract

In order to give consumers access to all the crucial details or the excitement of the films, video summarising seeks to create a high-quality text-based summary of the videos. In order to summarise a video, audio files must first be converted to text files, and then the video files. This whole Natural Language Processing's transformer architecture is used in conjunction with the process. Despite the fact that there have been numerous studies on text summarising, we propose our model, an extractive-video summarizer, which is based on cutting-edge pre-trained ML models and open-source libraries. The following regime is utilised by the extractive video summarizer: Creating a transdisciplinary collection of movies, extracting audio from video files, and creating text from audio recordings are just a few examples. the text summarization using (V)Entity extraction and extractive summarizers. Hindi and English, two widely spoken languages in India, were the main subjects of our research. To sum up, our model performs remarkably well and produces appropriate tags for videos.