Robot Learning Method Based On Video

Abstract

With the arrival of 5G era and the continuous development of computer technology and digital video technology, the amount of video data is increasing day by day. Therefore, there is a technical requirement in real life that can significantly shorten the video duration while retaining the main content of the original video. The emergence of video summarization method successfully meets this demand. It can reduce the time for people to browse videos and save a lot of storage space at the same time. Video summary is the process of automatically extracting key information from video and presenting it in the form of summary. It has been proved to be useful for both humans and machines because it can help them understand the main idea of video without watching the whole video. In this paper, we will focus on using deep learning technology for video summarization. Deep learning is an effective technology for many tasks, because it can learn complex functions by analyzing data samples. This also makes it an ideal summary technology, because video contains a lot of information needed. This paper studies various types of video, and proposes a robot learning method based on video summarization to improve the performance of video summarization algorithm and generate video summaries that interest users.