Research on the Method of Keras Library Combined with CNN Convolution to Classify Video

Abstract

With the popularization of the mobile Internet and the impact of the new crown epidemic, the time spent by netizens on the Internet is gradually increasing. Every minute, every major video website will add nearly 100 hours of video. How to help users quickly find their favorite content and achieve effective video push is the future development focus of deep learning technology. This paper mainly introduces the research direction of efficient video classification using Keras library combined with CNN.