The Application of Transfer Learning in Film and Television Works

摘要

Many personalized advertisement recommendation studies suffer from the problem of only certain tagged items can berecommended in video playback, which mean it can’t recommend more produces to users that they really like . It alsodoesn’t know the users really like at the source. Due to the large number of scene changes in different video, the userscan choose more items they like. This study attempts to adopt transfer knowledge to solve the problem of data volume toprovide users with a variety of options. Aiming at the image classification model of learning on big data set, this paperproposes a method to solve the problem of scene object recognition in TV program,such as movies,TV plays, and shortvideo, by transferring a pre-trained depth image classification model to a specific task. In a small training set, learninghigh-level representations on a small training set to produce a task-specific target model. Experiments on small data setsand real face sets collected by myself show that the transfer learning is effective and efficient. In the application of video,this study provides a theoretical basis for personalized click recommendation of video users.