Advancements in Communication and Systems
YouTube Transcript Summarizer using HuggingFace Transformer
Authors: Vijay Mane, Kshitija Dupare, Prashik Dupare, Tushar Ganvir and Rutvik Ghode
Publishing Date: 24-01-2024
ISBN: 978-81-955020-7-3
Abstract
In the vast online realm, where tons of videos pop up every day, our busy lifestyles make it tough to sit through long films. Complicating matters, many videos lack real content and instead use tricks like fake thumbnails and ads to catch attention. The massive growth in YouTube users adds to the flood of videos. To tackle this issue, our project focuses on giving users quick access to useful transcripts, saving them from wasting time on content that doesn’t offer much. We aim to speed up how users get knowledge by letting them easily read through video transcripts, avoiding the problems of misleading visuals and ads. Impressively, our project reached a high accuracy of 94.29% in creating these transcripts, especially for videos in English. Behind the scenes, we use advanced models for precise technical work. Looking ahead, we plan to improve the project by making it work with different languages and refining accuracy through smarter language understanding and pattern recognition. This keeps our project as a handy tool for smoothly handling the complexities of watching videos online.
Keywords
HuggingFace converter, Text Summarizer, Chrome Extension, Web API, Flask API.
Cite as
Vijay Mane, Kshitija Dupare, Prashik Dupare, Tushar Ganvir and Rutvik Ghode, "YouTube Transcript Summarizer using HuggingFace Transformer", In: Ashish Kumar Tripathi and Vivek Shrivastava (eds), Advancements in Communication and Systems, SCRS, India, 2024, pp. 145-153. https://doi.org/10.56155/978-81-955020-7-3-13