Abstract

When users know the information delivered within parts of a video, it is wasteful to watch those parts. Additionally, some videos may not be understood by a user that lacks certain background knowledge without viewing related content first. This disclosure describes techniques, implemented with user permission, to offer personalized options for users to skip known parts of a video and/or to splice in relevant supplementary content that adds content with relevant background information. With user permission, a personalized knowledge model of the user and a transcript of the original video the user wishes to watch are input to a large language model (LLM) along with a prompt to adjust the video. The LLM output includes an annotated video transcript indicating parts of the video to be skipped and places in the video where supplementary external content is to be inserted, along with identification of such content/ generated content. A user interface is provided to enable the user to skip known content and view supplementary content spliced in at appropriate places during video playback.

Creative Commons License

Creative Commons License
This work is licensed under a Creative Commons Attribution 4.0 License.

Share

COinS