Unattended Video Classifying System based on Transfer Leaning
Download PDF

DOI

10.26689/jera.v3i5.989

Submitted : 2019-11-20
Accepted : 2019-12-05
Published : 2019-12-20

Abstract

As the internet techniques advances, the demand of entertainment of the general public increases rapidly. A majority of short video applications like “TikTok” have appeared. The unbelievable growth of the number of the new videos and the increase of the average wage makes it almost impossible to identify the categories of the videos manually. In order to cater to the users’ preferences efficiently, detecting the content of the videos is inevitable. Fortunately, due to the great development on both the algorithm and hardware, the golden era of artificial intelligence is coming which makes it is possible to use computers to recognize the content of the videos. Comparing to the traditional method of content recognition, machine learning increases the efficiency with low cost, which is a better way to cope with the conflicts between the great increase of the number of the new videos uploaded by users and the lack of human resources.