SlidesScriber: An Efficient and Robust System for Automated Extraction of Presentation Slides from Teaching Videos

XJTU 2024 CSUC Submission3 Authors

26 Mar 2024 (modified: 03 Apr 2024)XJTU 2024 CSUC SubmissionEveryoneRevisionsBibTeXCC BY 4.0
Keywords: automated extraction, image processing, sequential outlier detection, content retrieval
Abstract: SlidesScriber presents a novel solution for the automated extraction of presentation slides from teaching videos in university recording platform. The system offers efficiency and robustness in identifying and extracting slides. Leveraging image processing and sequential outlier detection algorithms, SlidesScriber accurately samples video frames to detect and extract slides, facilitating efficient content retrieval. Through massive uses by students, the system demonstrates robust performance across various video recording environments, providing a valuable tool for learners in accessing and utilizing educational resources effectively.
Submission Number: 3